Monitoring bioreactors using principal component analysis: production of penicillin G acylase as a case study

Multivariate Analysis and Statistics in Pharmaceutical Process Research and Development

The application of statistics in pharmaceutical process research and development has evolved significantly over the past decades, motivated in part by the introduction of the Quality by Design paradigm, a landmark change in regulatory expectations for the level of scientific understanding associated with the manufacturing process. Today, statistical methods are increasingly applied to accelerate the characterization and optimization of new drugs created via numerous unit operations well known to the chemical engineering discipline. We offer here a review of the maturity in the implementation of design of experiment techniques, the increased incorporation of latent variable methods in process and material characterization, and the adoption of Bayesian methodology for process risk assessment.

A novel toolbox for E. coli lysis monitoring

The bacterium Escherichia coli is a well-studied recombinant host organism with a plethora of applications in biotechnology. Highly valuable biopharmaceuticals, such as antibody fragments and growth factors, are currently being produced in E. coli. However, the high metabolic burden during recombinant protein production can lead to cell death, consequent lysis, and undesired product loss. Thus, fast and precise analyzers to monitor E. coli bioprocesses and to retrieve key process information, such as the optimal time point of harvest, are needed. However, such reliable monitoring tools are still scarce to date. In this study, we cultivated an E. coli strain producing a recombinant single-chain antibody fragment in the cytoplasm. In bioreactor cultivations, we purposely triggered cell lysis by pH ramps. We developed a novel toolbox using UV chromatograms as fingerprints and chemometric techniques to monitor these lysis events and used flow cytometry (FCM) as reference method to quantify viability offline. Summarizing, we were able to show that a novel toolbox comprising HPLC chromatogram fingerprinting and data science tools allowed the identification of E. coli lysis in a fast and reliable manner. We are convinced that this toolbox will not only facilitate E. coli bioprocess monitoring but will also allow enhanced process control in the future.

Multivariate PAT solutions for biopharmaceutical cultivation: current progress and limitations

Increasingly elaborate and voluminous datasets are generated by the (bio)pharmaceutical industry and are a major challenge for application of PAT and QbD principles. Multivariate data analysis (MVDA) is required to delineate relevant process information from large multi-factorial and multi-collinear datasets. Here the key role of MVDA for industrial (bio)process data is discussed, with a focus on progress and limitations of MVDA as a PAT solution for biopharmaceutical cultivation processes. MVDA based models were proven useful and should be routinely implemented for bioprocesses. It is concluded that although the highest level of PAT with process control within its design space in real-time during manufacturing is not reached yet, MVDA will be central to reach this ultimate objective for cell cultivations.