Recovery & Purification
Characterization of a Recombinant Enzyme Downstream Process Using Modern DOE Techniques
Process characterization studies are a key component of a Quality by Design (QBD) approach, enabling improved process understanding, determination of critical process parameters (CPP), and definition of the process design space, ensuring that the process is capable of reproducibly generating safe and efficacious product for patients. While a classical design of experiments (DOE) approach is typically used for these studies, there are new statistical and data analytical techniques available that can improve experimental efficiency and better utilize the wealth of data generated from these studies. We demonstrate the potential for these new techniques with a case study of the downstream process characterization of a recombinant enzyme process. Of the four chromatography operations studied, one proved to be a particular challenge. During the initial multivariate screeni ng study, fully 25% of the experiments resulted in extremely low product recovery, indicating that the original process set-point was not robust. Because a new type of screening DOE, a definitive screening design (DSD), was used for this study, we were able to quickly and efficiently identify a new region of the design space for further characterization. Additional experiments were added in this region to support a response surface model, enabling the use of process simulation to identify a robust operating set-point and process design space. Utilization of these new statistical methods greatly improved the efficiency of the characterization of this complex chromatography operation. These techniques, especially the application process simulation for robust process engineering, can become a part of the standard toolbox for future QBD studies.