Digital Engineering for Higher Quality in Pharmaceutical Manufacturing

Quality by design (QbD) is an approach to ensure quality from the design stage by utilizing scientific evidence and quality risk management in the pharmaceutical development and manufacturing process. A pharmaceutical process was established based on a physical model that follows principles, and a design space was set up using the QbD method that utilizes the models. Conventionally, in pharmaceutical development and manufacturing, critical process parameters (CPPs) that affect critical quality attribute (CQA) are extracted by design of experiments (DoE) using orthogonal tables, and CQA are approximated by CPPs using a low-order formula. Using the model, CPPs can now be systematically determined through sensitivity analysis. In addition, the CPPs selected by sensitivity analysis may have parameters that cannot be experimented with because it is difficult to change the conditions in practice. This paper focuses on dealing with risk management issues in the anti-solvent crystallization process using seed crystals in small molecule pharmaceuticals. It is difficult to create and experiment with some crystal samples with varying seed crystal size, which affects the crystal size distribution. This paper reports on the use of a digital twin to obtain a design space in which the size of seed crystals can be varied to ensure the quality of the pharmaceutical product.