In vitro – in silico – in vivo drug absorption model development based on mechanistic gastrointestinal simulation and artificial neural networks: Nifedipine osmotic release tablets case study

Deciphering nifedipine in vivo delivery from modified release dosage forms: Identification of food effect

“…In vitro dissolution data can serve as markers for targeting the desired drug delivery profile in vivo. In this context, the usefulness of gastrointestinal simulation for the prediction of in vivo drug dissolution and absorption was evaluated using the previously developed and validated GastroPlus TM model (21). In vitro dissolution data of investigated dosage forms obtained in the present study were used as input values for GIS.…”

“…Physiologically based dissolution media simulating pre-and postprandial states in the stomach and small intestine, namely Fasted State Simulated Gastric Fluid (FaSSGF), Fed State Simulated Gastric Fluid (FeSSGF), Fasted State Simulated Intestinal Fluid (FaSSIF) and Fed State Simulated Intestinal Fluid (FeSSIF) were prepared as described by Jantratid et al (22). Samples were assayed for nifedipine UV spectrophotometrically or using high-performance liquid chromatography, as described previously (21). For FeSSGF with milk as a major component, 2 mL of isopropyl alcohol was added to a 3-mL sample, mixed well and the mixture was centrifuged at 3500 rpm for 5 min.…”

“…In order to take into account the potential food effect, drug absorption was simulated using the 'Human-Physiological-Fed ' mode in the GastroPlus settings. The gastrointestinal simulation (GIS) model development and validation were described previously (21).…”

In Vitro-In Vivo Correlation (IVIVC): From Current Achievements Towards the Future

“…Moreover, if the IVIVR transposes experimental results obtained in vitro to predict response-inducing properties, it becomes "in vitro-in vivo extrapolation" (IVIVE) (28). The first approach was made by Dowell et al, who used three types of artificial neural networks (feed forward, recurrent, jump connections, and general regression neural networks) to directly map the dissolution profile with pharmacokinetic observations (29). The resulting, so called, "input-output" associations were then externally assessed by calculating the correlation coefficient (R 2 ), mean prediction error, and mean absolute error.…”

From Heuristic to Mathematical Modeling of Drugs Dissolution Profiles: Application of Artificial Neural Networks and Genetic Programming

“…The artificial neural networks (ANNs), which are usually computer programs designed to simulate biological neural systems in both their functional activity and structure, are useful data analysis tools to handle such complex, nonlinear relationships. ANNs were utilized in multiple applications in various areas of science and technology [ 6 ] along with pharmaceutical sciences, including but not limited to development of nimodipine consisting floating tablets [ 7 ], minimisation of the capping tendency during tableting process [ 8 ], optimization of galenical dosage form technological process [ 9 ], prediction of dissolution from ketoprofen consisting solid dispersions [ 10 ], analysis of parameters of direct compression tableting [ 11 ], control of quality attributes of tablets manufactured by wet granulation [ 12 ], prediction of gentamicin [ 13 ], remifentanil [ 14 ] and aminoglycosides [ 15 ] serum concentrations, analysis of pharmacokinetic population data [ 16 ], evaluation of an in vitro-in vivo correlation for nebulizer delivery of salbutamol [ 17 ], oral verapamil products [ 18 ], sustained release paracetamol matrix tablet formulations [ 19 ], and nifedipine osmotic release tablets [ 20 ]. However they are also considered the so-called black-box models as it is not possible to provide analytical solution of their way of data processing; what stays in contradiction to the idea that the model purpose is to reveal information about the analyzed system.…”