Résumé : When considering oral administration, drug release from its pharmaceutical form and its dissolution into gastrointestinal fluids generally precedes absorption and systemic availability. The solubility-dissolution behaviour of a drug is frequently the rate-limiting step to absorption of drugs from the gastrointestinal tract (BCS class II drugs). Poor aqueous solubility has always been a very challenging obstacle as it is, together with membrane permeability, an essential factor in the limitation of a drug’s bioavailability following oral administration. Since an increasing number of newly developed drug candidates in pre-clinical development phases present poor water-solubility characteristics, there is a great need for formulation approaches to overcome this factor.

Out of the many ways to increase a product’s solubility/dissolution rate characteristics with the aim of enhancing its oral bioavailability, drug formulation as nanoparticles has received much-increased interest over the last decade. The hypothesis behind dissolution rate enhancement, considering drug particle size reduction to nanometer range, lies primarily in a much-increased effective surface area (Noyes-Whitney) presented by the resulting drug nanoparticles. Out of the various technologies available for drug particle size reduction to nanometer range, milling using high pressure homogenization is regarded as one of the simplest and most effective techniques. High pressure homogenization is a solvent-free process and is relatively rapid (time-saving). Furthermore, and most importantly, the scaling up of this technique is already established; processing capacities ranging from 3 l/h (e.g. EmulsiFlex C3®: minimum sample volume - 10 ml) to 1000 l/h (e.g. EmulsiFlex C1000®: minimum sample volume - 2 l).

Four model drugs were studied in this work. Nifedipine (NIF), an extensively studied poorly water-soluble drug in the literature, was used as the main model on which most of the development was done. In parallel to the work carried out on NIF, three UCB S.A. molecules currently under development were also studied as poorly water-soluble drugs: these being ucb-35440-3, UCB-A and UCB-B (salt of UCB-A). These three UCB S.A. model drugs are, contrarily to NIF, predicted highly dosed drugs and are weak bases, and thus present pH-dependent solubility profiles, which allowed us to investigate model drugs with different profiles.

Firstly, investigations regarding appropriate formulation development (stabilizer (surfactant) selection) and appropriate high pressure homogenization operating conditions (pre-milling cycles, influence of the number of high pressure homogenizing cycles, influence of homogenizing pressure, influence of sample temperature) were made. It has been shown, through this development, for the four studied model drugs, that high pressure homogenization is an appropriate technique for reducing drug particle size to nanometer range (NIF  290 nm, ucb-35440-3  180 nm, UCB-A  350 nm and UCB-B  250 nm). Investigations regarding water-removal from the nanosuspensions obtained and most importantly regarding the redispersion characteristics of the retrieved powders (i.e. nanoparticles) were then carried out. In that regard, it has been shown that the presence of carriers in the formulation is essential for limiting nanoparticles agglomeration during the water-removal operation.

Drug crystalline state characterizations before and following particle size reduction were then carried out on the three studied model drugs, mainly through DSC and PXRD studies. In fact, one of the advantages of this particle size reduction approach (using high pressure homogenization), versus other frequently studied solubility/dissolution rate enhancement technologies (e.g. such as solid dispersions), is that original crystalline state shall not be altered in such a way that the achieved increased solubility and dissolution rate characteristics do not rely on the presence of the amorphous form of the drug; this furthermore implying a greater time-stability of the developed formulations. Through the data obtained, it has been shown that original drug crystalline state seems to be unaltered following particle size reduction.

In vitro solubility and dissolution characteristics were then evaluated on the formulations developed in order to verify the posed hypothesis regarding effective surface area increase. It has been shown through these studies that drug solubility and most importantly drug dissolution rate can be significantly enhanced for nanoparticulate systems (verified for NIF, ucb-35440-3, UCB-A and UCB-B). For example, solubility was enhanced from 26 µg/ml vs. 19.5 µg/ml for NIF nanoparticles and the dissolution characteristics showed that 100% of the tested dose (equivalent to 10 mg NIF) was already dissolved following 10 min vs. less than 5% for un-milled NIF. Following these very interesting and promising results, and preliminary to the in vivo pharmacokinetic studies carried out, in vitro permeation studies (apical to basolateral transfer studies) across intestinal cell models (Caco-2 and HT29-5M21 cultures and co-cultures) were carried out. This evaluation was only carried out using NIF as a model drug and showed a 6-fold increase in the permeation rate for NIF nanoparticles. The influence of chitosan (permeability enhancer/bioadhesive polymer) in the NIF nanoparticle formulation with regard to in vitro NIF permeation rate was also evaluated.

In vivo pharmacokinetic studies in rats were conducted using NIF and ucb-35440-3 as model drugs. The very different profiles of these two model drugs allowed us to retrieve interesting information regarding the in vivo behaviour of the developed formulations. As expected from the in vitro (i.e. solubility/dissolution/permeation) studies and results obtained for NIF, an increased extent of exposure could be observed for NIF nanoparticles versus un-milled NIF; the difference being more pronounced when the formulations were orally administered into capsules (2.5-fold increase in extent of exposure and 6-fold increase in Cmax). For ucb-35440-3, a poorly water-soluble weak base with a reported significant food effect considering oral bioavailability, an increased extent of exposure for nanoparticles, versus the un-milled drug, could only be observed in fasted state (4-fold increase in extent of exposure and 2.7-fold increase in Cmax). These different, diet-relative observations allowed us to put forward some limitations and precautions (considering poorly water-soluble weak bases) relative to the possibility of drug reprecipitation following stomach’s exiting, particularly if dissolution in the stomach is quite fast (e.g. nanoparticulate systems).

In parallel to the in vivo pharmacokinetic evaluation of NIF nanoparticles, evaluation of the antihypertensive effect of the systems developed following oral administration, using spontaneously hypertensive rats, was also carried out and compared to un-millled NIF. The results obtained showed a significant drop in systolic blood pressure for NIF nanoparticles (32% reduction of initial SBP following 30 min vs. 1% for un-milled NIF) and nicely complemented the in vitro and in vivo results obtained for NIF nanoparticles.

Finally, a stability study of the optimized NIF nanoparticle formulation was carried out with respect to reported ICH conditions (25°C/60% RH; 30°C/65% RH; 40°C/75% RH). The results showed that the studied NIF nanoparticle formulation retains all its original characteristics (dissolution, crystalline state, redispersion characteristics); this being verified over time (12 months) and for each of the three storage conditions studied.