Abstract

Using DoE under Real-Life Restrictions

Philippe Solot, Chief Executive Officer, AICOS Technologies AG

Statistical design of experiments (“DoE”) is one of the main tools ensuring quality from the early stages of development, and thus more and more required by the regulatory agencies. It can be used for identifying critical process parameters (CPP) and material attributes (CMA), in order to find optimum formulations and process parameter settings, to facilitate transfer and scale-up, to ensure robustness, as well as for reducing costs. Dr. Philippe will illustrate some of these points with case studies, but then concentrate on the central example of the development of a modified-release tablet, which was conducted at the university of applied sciences of Ostwestfalen-Lippe, Germany, in cooperation with AICOS Technologies, Switzerland. There, the aim was to obtain a tablet with a specific dissolution profile. Several practical issues were complicating the project: some process parameters could not be varied independently, specifying increments for the factors was necessary, several response variables had to be combined, and suddenly cracks appeared in the surface of the tablet. Nevertheless, with the structured QbD approach used, a valid design space could be defined quite smoothly