A number of therapeutic proteins are used for long in clinical practice. These include for example insulin, calcitonine, growth hormone, and parathyroid hormone for the treatment of various systemic disorders, as well as protein antigens in vaccine formulations. Due to the recent developments in biochemical engineering and in the comprehension of the physiopathology of many diseases, peptides and proteins are expected to become a drug class of increasing importance. Recently, novel biological drugs have for example been developed such as monoclonal antibodies, antibody fragments, soluble receptors, and receptor agonists or antagonists. These are mainly used for the treatment of auto-immune and inflammatory diseases (asthma, rheumatoid arthritis) and for the treatment of cancers. However, a major drawback of these biomolecules is the need to use parenteral administration. This is mainly due to the harsh pH conditions that proteins undergo by oral administration, leading to various physico-chemical degradations and loss of biological activity. <p><p>Pulmonary delivery of these proteins could constitute an alternative to parenteral delivery. Due to the very high surface area of the lungs, the low thickness of the alveolar epithelium and the high level of lung vascularisation, pulmonary administration can indeed provide fast systemic absorption of drugs, while avoiding hepatic first pass metabolism. On the other hand, drugs for local treatment can also be administered directly into the lung, which allows delivering high doses while limiting systemic side effects. Nevertheless, administration of drugs to the lungs requires some challenges to be taken up. It is indeed necessary to provide the drug as very small solid or liquid microparticles (1-5 µm) in order to reach the lungs. For solid microparticles, it is also needed to overcome the very high inter-particle interactions by using appropriate formulation strategies and by including deaggregation mechanisms in the inhalation device. Other issues are more specifically related to the pulmonary administration of proteins. These can indeed undergo physico-chemical degradations during processing, administration, and/or storage. Moreover, if systemic action is required, proteins will often need addition of an absorption enhancer to cross the alveolar epithelium because of their large molecular weight and hydrophilicity. <p><p>In this work, we developed formulations for pulmonary delivery of proteins using two model proteins. Insulin (5.8 kDa) was chosen as a model of small protein. It is also an application of systemic pulmonary delivery. On the other hand, an anti-IL13 monoclonal antibody fragment (54 kDa) was used as a model of larger protein. This molecule is currently in development for the treatment of asthma and provided an application for local pulmonary delivery. The formulation strategy was to produce dry powders using a combination of micronisation techniques (high speed and high pressure homogenisations), drying techniques (spray-drying, freeze-drying), and addition of lipid excipients. These lipid excipients were added as a coating around the protein particles and were expected to prevent protein degradations during processing and/or storage, essentially by avoiding contact with water. It could also improve the aerodynamic properties of the powders by modification of the surface properties of the particles and/or limitation of the capillary forces.<p><p>First, we evaluated insulin lipid-coated formulations and formulations without excipients, produced using high pressure homogenisation and spray-drying. In the case of lipid-coated formulations, a physiological lipid composition based on a mixture of cholesterol and phospholipids was used. We were able to obtain good aerodynamic features for the different formulations tested, with fine particle fractions between 46% and 63% versus 11% for raw insulin powder. These are high FPF values in comparison with those obtained for other protein formulations for inhalation currently under development, which often have an in vitro deposition of around 30%. Insulin presented a good stability in the dry state, even when no lipid coating was added.<p>The presence of a lipid coating of up to 30% (w/w) did not significantly improve the aerodynamic behaviour of the powders, but the coated formulations exhibited decreased residual moisture content after 3-month storage, which should be of interest for the long-term stability of the formulations. <p><p>In a second step, two of the developed insulin formulations were evaluated in a clinical study to determine whether the formulations give high deep lung deposition in vivo, and how insulin is absorbed into the systemic blood stream. This pharmaco-scintigraphic trial was performed on twelve type 1 diabetic patients using an uncoated formulation and a formulation coated with 20% (w/w) of lipids. The two formulations showed interesting features, with pharmacokinetic profiles that mimic the natural insulin secretion pattern. Bioavailability was within the ranges of two of the three dry powder insulins that have reached phase III clinical development. However, the formulation with a lipid coating exhibited a lower lung deposition in comparison with the uncoated formulation, which was not expected from the previous in vitro results. Additional in vitro experiments indicated that this lower performance was related to a decrease in the disaggregation efficiency of the powder at a sub-optimal inhalation flow-rate. An extensive training of the patients to the inhalation procedure could therefore improve the lung deposition of the coated formulation.<p><p>Finally, we developed and evaluated dry powder formulations of the anti-IL13 antibody fragment. These were produced using, successively, freeze-drying, high pressure homogenisation (HPH), and spray-drying. The influence of different types and concentrations of stabilising excipients was evaluated for each production step. Due to its more elaborated structure, the antibody fragment was found to be more sensitive than insulin to physico-chemical degradation, particularly during the HPH process, which led to different types of degradation products. These could partly be avoided by adding 50% sucrose during freeze-drying and 10% Na glycocholate or palmitic acid in the liquid phase during HPH (dispersing agents). However, the presence of a small fraction of insoluble aggregates could not be fully avoided. Further spray-drying of the suspensions in the presence of 10% Na glycocholate or palmitic acid led to the formation of a hydrophilic or hydrophobic coating around the particles, respectively. Na glycocholate was found to be particularly effective in protecting the antibody during spray-drying, which was found to be at least partly related to its ability to inhibit sucrose recrystallisation. However, the best formulation still presented a small fraction of insoluble aggregates (6%). The aerodynamic evaluation of the formulations showed FPFs that were compatible with lung deposition, with the formulation containing Na glycocholate presenting the highest FPF (42%). The formulation coated with palmitic acid presented a slightly lower FPF (35%). The aerodynamic properties of this formulation remained unchanged at a sub-optimal inspiratory flow rate, to the contrary of what was observed for the insulin formulation coated with 20% (w/w) cholesterol and phospholipids. Palmitic acid could therefore be of interest as a hydrophobic coating material, and provide long-term stability of protein drugs. <p>The work performed with the insulin and anti-IL13 molecules provided the proof-of-concept that it was possible to obtain dry powder protein formulations with appropriate aerodynamic properties and good overall physico-chemical stability, using simple production techniques and few selected excipients. The formulation strategy presented in this work could therefore be of interest for the future development of inhaled proteins for local or systemic applications. <p> / Doctorat en sciences pharmaceutiques / info:eu-repo/semantics/nonPublished