In silico and in vitro models in pharmacokinetic studies

Abstract

One of the aims of the thesis was to design and realize an in vitro device able to reproduce the gastrointestinal behavior. To reproduce the temperature and pH history an USP apparatus II coupled with a control system was used. The temperature was kept constant using the USP apparatus, a pH probe was inserted in the dissolution medium to measure the pH. The measured pH was compared (by a software) with a set point. Proportionally at the mean error, a quantity of an acidic or basic solution was inserted, by pumps, in the dissolution medium adjusting the pH at the desired value. Using the real pH history of the gastrointestinal tract, which provide a decrease in the pH value from 4.8 to about 2.0 during the first two hours of dissolution, and then an increase to 6.8, the release pattern from tablets was evaluated. The release patterns of these tablets obtained with the new device were compared with those obtained using the conventional method (which provides a pH 1 during the first two hours of dissolution, and then the neutralization at pH 6.8) and it was found that the drug released during the first two hours was higher in the case in which the real pH history was reproduced. This is due to the fact that the higher pH in the first stage damages the coating of the tablet. Once the chemical and thermal conditions were reproduced, the reproduction of the transport across the intestinal membrane was faced. An high throughput device which is able to reproduce continuously the exchange between the compartments has been necessary. The USP apparatus was equipped with a device composed by an hollow filter (which simulate the intestinal wall) and two pumps for the fluids simulating the intestinal content and the circulatory system surrounding the gastrointestinal tract content. The fluids enter in contact in the filter and the fluid rich in drug content (that simulates the intestinal content) gives the drug to the fluid poor in drug (simulating the blood content). The release patterns obtained by the use of this device were studied and compared with those obtained following the conventional dissolution method. Moreover these release patterns obtained using the real pH evolution were coupled with the effect of mass exchange and compared with those obtained using the conventional methods. The results showed that the effect of the real history of pH is higher in the first stage of dissolution, than the effect of the mass exchange is dominant. The reproduction of the mechanical history of the stomach is than faced. The peristaltic waves were reproduced using a lattice bag (elastic and compressible) connected to a camshaft which, with its rotation ensured the contraction of the bag. The bag was shrunk by connectors and the right position was ensured by guides. Changing the rotation speed of the shaft, the frequency of the contractions could be adjusted. The release pattern of a commercial tablet in the new device was evaluated and compared with the conventional one. The results showed that the non-perfect mixing of the stomach was satisfactory reproduced and this lead to a release pattern completely different. Moreover, the effect of the frequency of the contractions on the release pattern was evaluated. Second, but not secondary, aim of the thesis was to develop an in silico model (physiologically based) which is able to simulate the plasma concentration of drugs. The model is composed by seven compartments, which simulate the human organ, tissue, or a group of them. The compartments are interconnected between them and seven differential equations (with their initial conditions) describe their behavior. Once the parameter are obtained (by fitting or in literature), using an in vitro release pattern, the model is able to simulate the concentrations in all the compartments, including the plasma compartment. The plasma concentration are simulated both in the case in which the new release pattern (with the real pH history) is used as input, and the case in which the conventional one is used. The results show that in the real case the plasma concentration is very different both in value and in shape than the expected. The model then was used to simulate the fate of several molecules simultaneously in the human body (i.e. if a racemic mixture is administered or if the drug is metabolized to another molecule). The system of differential equations is expanded to describe the fate of each molecule. Then, the physiological parameters, such as gender and age, were integrated in the model; in this way, the dependence of the model parameter on the physiological parameter was evaluated. Finally, the gastrointestinal concentration simulated with the in silico model was successfully compared with the drug concentration measured with the in vitro model. It could be concluded that the combined approach which uses the in vitro and the in silico models is a powerful tool in the pharmacokinetic studies. [edited by Author]