(M0930-02-13) Predicting Food Effects of Commercial Products Using the Dynamic Gastric Model (DGM)

Purpose: Complex in vitro models of the upper gastro-intestinal (GI) tract that mimic the dynamic GI environment (in terms of hydrodynamics and fluid composition) are potentially of high value for the development and understanding of new oral pharmaceutical products as they may be capable of providing a similar or better predictive ability than animal models. The dynamic gastric model (DGM) is one such model that simulate the hydrodynamics, emptying and pH profile of the fed or fasted stomach. The DGM is capable of accommodating a glass of water, as well as a full meal (1). Thus, the DGM can be used to evaluate the in vivo behavior of a dosage form in both the fasted and fed state. The aim of the current study was to evaluate if the DGM in combination with a duodenal module, is able to predict the food effect of marketed products containing diclofenac potassium (DIC), guaifenesin (GUA), cinnarizine (CIN) and paracetamol (PAR). Simple convolution of the duodenal dissolution data was performed in order to predict in vivo pharmacokinetic (PK) data.

Methods: Fasted state DGM experiments: The formulation was administered along with 240 mL of tap water (ambient temperature) to the DGM, which contained 20 mL of gastric acid solution. During the experiment, gastric acid solution and gastric enzyme solution were added at a dynamic rate depending on pH and stomach content, respectively, to reach pH~2. The processing time was 29 min over which 7 samples of approx. 40 mL were ejected every 4 min. Immediately after ejection, a sample of 1 mL was withdrawn. This sample was centrifuged at 10,000 rpm for 10 min, and the supernatant diluted with an organic solvent and analyzed for drug concentration using HPLC-UV. The remaining sample was transferred to a 600 mL thermostated jacketed glass vessel equilibrated to 37 °C under stirring, along with 5 mL of concentrated (x9) fasted state simulated intestinal fluid (FaSSIF). Throughout the experiment, the pH was continuously adjusted to 6.5 through addition of 1M NaOH using a 902 Titrando from Metrohm (Herisau, Switzerland). Prior to addition of the following DGM sample, a sample of 1 mL was withdrawn from the vessel. Samples were treated as described previously. After addition of the last DGM sample, the intestinal digestion process was continued for 270 min with sampling every 15 min. Fed state DGM experiments: A high-fat FDA breakfast (1000 kcal, 50% fat) was prepared, chewed by a human volunteer and spit into the DGM, which contained 20 mL of gastric solution. After 30 min of digestion (in line with FDA fed state protocol), the formulation was administered to the DGM along with 240 mL of tap water (ambient temperature). During the experiment, gastric acid solution and gastric enzyme solution were added at a dynamic rate depending on pH and stomach content, respectively, to reach pH~2. The processing time was 257 min over which 17 samples of approx. 70 mL were ejected every 16 min. Immediately after ejection, the weight and pH of the samples were measured and a sample of 1 mL was withdrawn. Samples were treated as described previously. The remaining sample was transferred to a 2000 mL thermostated jacketed glass vessel equilibrated to 37 °C under stirring, along with 10 mL of concentrated (x8) fed state simulated intestinal fluid (FeSSIF). Throughout the experiment, the pH was continuously adjusted to 5.8 through addition of 1M NaOH using a 902 Titrando from Metrohm (Herisau, Switzerland). Prior to addition of the following DGM sample and 10 ml concentrated FeSSIF, a sample of 1 ml was withdrawn from the vessel. Samples were treated as described previously. After addition of the last DGM sample, the intestinal digestion process was continued for 80 min with sampling every 16 min.

Results: An overview of the DGM-duodenal module will be given along with data on four different marketed drug products. An example of the dissolution profiles (from CIN) of the samples from the DGM and the duodenal module are shown in Figure 1 along with the predicted plasma profile based on convolution and PK data. As can be seen, despite the fast release of drug in the fasted state stomach compartment, the drug quickly precipitates in the intestinal compartment. In contrast, in fed state stomach compartment there is a steady release of drug over time, which does not precipitate in the intestinal compartment. In turn this drug precipitation in the fasted state is causes the positive food effect shown in the clinical study and as predicted by the PK prediction model.

Conclusion: The DGM-duodenal model combined with simple convolution is able to predict the human plasma concentration-time profile of four different marketed products with great precision in both fasted and fed state conditions. This suggests that the DGM is can predict food-effects of drug products, which provide valuable information and could prevent clinical failure.

References: 1. Wickham, M.J.S., Faulks, R.M., Mann, J., & Mandalari, G. (2010). The design, operation, and application of a dynamic gastric model. Dissolution Technology, 19(3), 15-22.
2. Qureshi, S.A. (2010). In vitro-in vivo correlation (IVIVC) and determining drug concentrations in blood from dissolution testing – a simple and practical approach. The Open Drug Delivery Journal, 4(1), 38-47.
3. Berlin, M., Przyklenk, K. H., Richtberg, A., Baumann, W., & Dressman, J. B. (2014). Prediction of oral absorption of cinnarizine–a highly supersaturating poorly soluble weak base with borderline permeability. European Journal of Pharmaceutics and Biopharmaceutics, 88(3), 795-806.

Figure 1: Example of gastric dissolution, duodenal dissolution and plasma profiles during fasted state (blue) and fed state (red) conditions in the DGM-duodenal model (solid line) and in vivo in humans (dashed line) for CIN (3).