Predicting the pharmacokinetics and food effect of oral drug products using the dynamic gastrointestinal model (DGM)

Abstract

The pharmacokinetics (PK) of oral drug compounds are often significantly altered by food intake and evaluating this effect, as required by regulatory agencies, typically involves costly and time-consuming clinical trials. This study used the Dynamic Gastrointestinal Model (DGM), an advanced in vitro system simulating both biochemical and mechanical aspects of the human upper gastrointestinal tract, to predict plasma concentration–time profiles (PK profiles) and food effect of three immediate release oral drug products. The drug products, containing cinnarizine (CIN), diclofenac potassium (DIC) or paracetamol (PAR), were processed in the DGM mimicking the fasted and fed state clinical protocols and the resulting intestinal drug dissolution profiles were modelled (by convolution) to achieve the predicted PK profiles. The predicted PK profiles in both the fasted and fed state were in accordance with the observations in clinical trials, capturing both the positive food effect for CIN and the negative food effects for DIC and PAR. These findings demonstrate the ability of the DGM to provide insights into the PK performance and food effect of oral drug products.

Introduction

The rate and extent of drug absorption from oral dosage forms may vary significantly depending on fasted or fed state conditions, which can have therapeutic implications [1]. In fact, according to a review by O’Shea et al., more than 40 % of drug products approved by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA) between 2010 and 2019 display significant food effect (defined as the 90 % confidence intervals of AUC and/or Cmax between fed and fasted treatments fail to meet the 80–125 % bioequivalence limits) [2]. Therefore, as a part of new drug applications, the FDA requires the assessment of influence of food intake on the pharmacokinetic (PK) performance of an oral drug product (i.e. food effect) evaluated in human volunteers preferably upon intake of a high-calorie and high-fat meal (FDA meal), 30 min prior to drug administration [3]. However, since clinical trials are costly and time-consuming there is a demand for preclinical models, in silico models (such as physiologically based pharmacokinetic models) or in vitro models that can predict the food effect and/or guide the design of the clinical trials, thereby minimize costs and mitigate failure. Animal models have historically been preferred for this purpose as they generate in vivo data, but they are also associated with increased ethical concerns and rarely predictive of food effect in humans due to significant physiological differences [4]. In contrast, advanced in vitro models have no ethical constraints, but they lack biological feedback mechanisms, which can make it difficult to simulate in vivo data. However, in vitro conditions can be controlled, and therefore, the predictive quality of these in vitro models depend on the accuracy of the simulated gastrointestinal (GI) conditions [5,6]. The role of the stomach as a pivotal factor in this regard has recently gained growing attention as reflected in the increasing number of in vitro models aiming at simulating the human gastric physiology [7,8]. Of these, the improved human gastric simulator (HGS) [9], TNO intestinal model advanced gastric compartment (TIMagc) [6], biorelevant gastrointestinal transfer (BioGIT) system [10], gastric digestion simulator (GDS) [11], dynamic in vitro human stomach (DIVHS) [12] and the dynamic gastrointestinal model (DGM) [[13], [14], [15]] are amongst the more complex and physiological relevant in vitro models developed [7,8].

The DGM is a computer-controlled compartmental model designed to accurately simulate the physical/mechanical and biochemical processes that occur within the human upper GI tract. The gastric compartment of the DGMis owned by Plant Bioscience Ltd (Norwich, UK) and licensed to Bioneer A/S (Copenhagen, Denmark) who refined the intestinal compartment of the DGM to enable pharmaceutical testing. The DGM simulates the dynamic conditions of the stomach, including dynamic acid and enzyme addition, as well as gastric motility and mixing. Like the human stomach, the gastric compartment of the DGM consists of two functional regions; the body and the antrum [16]. The body of the DGM serves as a reservoir for the ingested food and includes addition of acid and enzymes along the sides of the walls while performing gentle massaging movements. The antrum of the DGM simulates the in vivo peristalsis and hydrodynamics, thereby mixing the food bolus and agitating the drug product [13,17]. The antral shear/forces were calibrated towards human in vivo data by use of agar gel beads of various fracture strengths and are mimicked by the sliding of a piston within a barrel, which forces the material through an elastic annulus [13]. The DGM is able to simulate any relevant gastric condition, such as increased stomach pH due to intake of acid reducing agents and/or ingestion of a chewed meal (without the need for homogenization or decreasing the volume), which provide valuable insights into the effects of food on drug product behavior. The digesting food bolus is ejected from the antrum of the DGM in a manner simulating human gastric emptying. Finally, the samples are incubated in the intestinal compartment of the DGM, which simulates the dynamic digestion in the small intestine through addition of bile, phospholipids, intestinal (pancreatic) enzymes and pH control.

The aim of the present study was to evaluate the food effect on the rate and extent of drug absorption demonstrated in clinical trials in the literature, by simulating the fasted and fed clinical protocols in the DGM. Three marketed products Sepan® (25 mg cinnarizine), Cataflam® (50 mg diclofenac K), and Panadol® (500 mg paracetamol) were used as model drug products. These drug products cover a representative range of different drug physicochemical properties such as charge at physiological pH and aqueous solubility, as well as displaying negative, positive and neutral food effects. The DGM intestinal dissolution data were modelled into pharmacokinetic (PK) profiles through a convolution approach and compared with the clinical PK profiles available in the literature [[18], [19], [20]].

Read more here

Materials

Cinnarizine, diclofenac, paracetamol, sodium chloride, potassium chloride, acetic acid glacial, calcium chloride, trifluoroacetic acid, sodium dihydrogen phosphate, hydrochloric acid, sodium hydroxide, pepsin from porcine pancreas, pancreatin from porcine pancreas and lipase R Oryzae were sourced by Sigma-Aldrich (St. Louis, MO, USA). Instant SIF powder was purchased from Biorelevant (London, United Kingdom). The drug products Sepan® 25 mg (Janssen, Belgium), Cataflam® 50 mg (Novartis.

Matthias Manne Knopp, Jacob Rune Jørgensen, Laila Tognarelli Hansen, Anette Müllertz, Predicting the pharmacokinetics and food effect of oral drug products using the dynamic gastrointestinal model (DGM), European Journal of Pharmaceutics and Biopharmaceutics, 2025, 114723, ISSN 0939-6411, https://doi.org/10.1016/j.ejpb.2025.114723.

Read also our introduction article on Orally Disintegrating Tablets (ODTs) here: