Development and optimization by experimental design of an in vitro method for prediction of drug buccal absorption

Drug administration via the buccal route is receiving growing attention, due to the advantages it presents with respect to the classic oral route. Moreover, the continuous advances in drug delivery technology are progressively increasing the variety of drugs, including biological agents, that can be delivered across the oral mucosa for systemic diseases treatment as an alternative to parenteral administration. Therefore, considering the great potential applications of this administration route, it is important to have a reliable in vitro method available for an easy and rapid evaluation of permeability through the oral mucosa of possible candidate drugs. Buccal drug delivery is commonly evaluated in vitro by diffusion studies across freshly excised animal mucosa, or using oral epithelial cells grown on filters. However, both methods present several drawbacks (difficult handling and maintenance of fresh excised tissue, high biological variability and consequent poor reproducibility, the first one; long cell growth cycles, risks of microbial contamination, high costs and rather elevated inter-experimental and inter-laboratory variability, the second one) which limit their use for a high throughput screening approach. In vitro methods based on the use of artificial membranes could represent an interesting alternative to the above methods, due to their greater simplicity, lower cost, shorter times and better reproducibility. In this light, the aim of this work was to develop an in vitro method for drug buccal absorption assessment, based on the use of a special artificial membrane purposely optimized. Naproxen was selected as reference model drug in this preliminary study. Its apparent permeation coefficient (Papp) through freshly excised pig buccal mucosa, determined using a suitably modified Sartorius apparatus, was considered as the target value to be obtained with the artificial membrane. An experimental design methodology was used to optimize the composition of the lipid mixture used for support impregnation. An initial screening phase provided indication on Papp variation trend as a function of the kind of phospholipid component, allowing selection of the most suitable one. The following response surface study, performed by a mixture design, enabled to define a design space where each combination of the lipid mixture components fulfilled the desired Papp target value within the prefixed confidence interval with a failure risk < 1%. Further studies with other model drugs have been planned to confirm the actual predictive ability of the method, which could offer a useful tool for a quick and effective screening in the early stages of drug discovery and/or in preformulation studies.