Evaluation of lipid-based formulations of poorly water-soluble drugs in the gastro-intestinal tract using in vitro tests

Novel active pharmaceutical ingredients are often poorly water-soluble. Such compounds may only partially dissolve or may precipitate during intestinal passage, potentially leading to incomplete drug absorption. Despite the importance of the process, the underlying in vivo as well as in vitro drug-precipitation mechanisms remain poorly understood. Several formulation principles, including lipid-based formulations, have been introduced to prevent drug precipitation in the gastro-intestinal tract. However, in vitro performance testing of these formulations is a topic of ongoing scientific discussions. Reliable in vitro tests as well as suitable monitoring tools to better analyze in vitro solubilization, precipitation, as well as lipolysis processes in the gastro-intestinal tract are required.
In the present thesis, dispersion, dissolution, precipitation, and lipolysis processes are discussed. We compared the results obtained with a paddle apparatus with those from a physiologically motivated flow-through cell taking lipolysis into consideration, using lipid-based formulations of a weakly acidic drug (Biopharmaceutics Classification System Class II). We tested pure indomethacin and the drug-containing self-microemulsifying drug delivery system (SMEDDS) using pure aqueous buffers and biorelevant media. The results of these dispersion/precipitation tests showed generally increased solubility of indomethacin in the SMEDDS compared with the solubility of the pure drug. One of the SMEDDS was superior compared to the others regarding the solubilizing capacity. This was demonstrated only in the flow-through test and dispersion in hydrochloric acid (0.1?N?HCl). However, these results must be interpreted in the light of the lipolysis test showing that the observed differences in solubilization were not based on lipolysis. We concluded that suitable characterization of SMEDDS involving an acidic drug should include a physiologically motivated flow-through test or dispersion/precipitation test in acidic environment, together with a lipolysis test.
We studied the effects of polysorbate?80 (PS80) on fenofibrate precipitation in the simulated intestinal medium using focused beam reflectance measurement (FBRM). We dissolved three different quantities of fenofibrate in six different mixtures of PS80 and ethanol (EtOH). After adding these formulations to biorelevant media, we evaluated the effects of micelles in the simulated medium in combination with PS80 on fenofibrate solubility and precipitation. Endogenous micelles in combination with PS80 micelles enhanced drug solubility and therefore reduced supersaturation. Compared to pure water, micelles of biorelevant media accelerated drug-precipitation kinetics. Addition of increasing amounts of PS80 to the biorelevant media prolonged nucleation time slightly and reduced the number of particles. We successfully introduced FBRM as a monitoring tool in biorelevant media.
In another drug precipitation test, we simulated the transfer from the stomach to the intestine using simulated gastric and intestinal media. We used online dynamic image analysis and inline Raman spectroscopy. Further, we analyzed concentration profiles of the model drug dipyridamole in the simulated intestine by high-performance liquid chromatography (HPLC), and we developed a kinetic nucleation and growth model that was fitted to the experimental data. Dynamic image analysis revealed a complex structure of the precipitated dipyridamole particles. These precipitated upon transfer to the intestinal medium and were described as star-like crystals or aggregates of elongated primary particles. Furthermore, Raman spectroscopy allowed the monitoring of precipitation over time. By fitting the model to the data, nucleation and growth exponents were obtained. These were consistent with data published in the literature and provided perfect agreement between the model and data.
The last part of the work described in this thesis focused on in vitro lipolysis of lipid-based drug delivery systems. Dispersion and digestion processes mainly govern the fate of such systems. We studied concentration effects of six poorly water-soluble drugs on in vitro lipolysis rate of medium-chain triglycerides (MCT), and we compared the results with drug effects on oil viscosity and surface tension. First, we characterized the drugs by molecular modeling and determined an apparent in vitro lipolysis rate in biorelevant medium by potentiometric titration.
The different drugs exhibited varying effects on oil viscosity and surface tension. However, all drugs significantly lowered the apparent lipolysis rate of the oil. This effect was very similar among the different compounds and did not correlate with the effects on oil viscosity and surface tension. Orlistat was the exception in that it practically blocked lipolysis by direct inhibition. The other drugs affected lipolysis kinetics most likely by different mechanism(s). In the light of the obtained results, drug effect on oil viscosity or surface tension appeared to play a minor role in reducing lipolysis rate. The lipolysis kinetics were not affected by the drug load, which was deemed advantageous from a pharmaceutical viewpoint. Different dose strengths are therefore not assumed to alter lipolysis kinetics, which is beneficial for limiting the variability of in vivo drug release.
Moreover, we studied the digestibility of 10 excipients often used in lipid-based drug delivery systems. We introduced a mathematical model to describe in vitro lipolysis kinetics, and we defined the relative half-lipolysis time that was independent of the set-up of the lipolysis test using Miglyol?812 as the reference excipient. The results indicated two classes of excipients. Some additives were partially hydrolyzed, while others displayed complete lipolysis. For the latter class, we used the lipolysis extent X as a function of time in a simplified mathematical model that provided a good first approximation of initial lipolysis kinetics. The relative half-lipolysis time was obtained from the model with Miglyol?812 as the reference and seemed to be a promising tool for comparing results of in vitro tests employing different experimental conditions.
In conclusion, the analytical tools and mathematical models provided new insights into in vitro solubilization, precipitation, as well as lipolysis in the gastro-intestinal tract. A more complete understanding already at an early stage of drug development allows the formation of new, much more efficient lipid-based drug delivery systems that minimize drug precipitation.