Predictive chemometrics modelling for enhanced robustness of integrated continuous biomanufacturing processes

Lead Research Organisation: University College London
Department Name: Biochemical Engineering


The proposed PhD project will address the challenge of developing robust continuous manufacturing processes for new modalities of biopharmaceutical proteins such as bispecific antibodies. Continuous bioprocessing is experiencing a resurgence within the sector. It has the potential to enhance product quality and radically reduce capacity and facility footprint requirements for manufacture. Furthermore, continuous processing is considered by some to be better able to address market demand uncertainties and minimise technology transfer risks. A rigorous approach to assessing the impact of introducing products adopting continuous bioprocessing on product quality, process robustness and portfolio management is required to determine the best route to commercialisation of new modalities. The department of Biochemical Engineering at UCL is well-placed to address this challenge as it has pioneered the development of decisional tools comprising bioprocess optimisation and chemometrics. These tools will enable ultrascale-down studies of integrated unit operations to better mimic large scale continuous bioprocesses, the enhanced understanding of which will address effective bioprocess design, portfolio management and capacity planning decisions.
This project aims to create experimental and predictive modelling tools to assess and optimise novel ideas for end-to-end continuous bioprocessing of biopharmaceuticals. The potential of continuous processing to maintain and control product quality of biologics will be evaluated.
MedImmune is evaluating the potential of integrated continuous bioprocessing at bench scale with ATF perfusion culture linked to multi-column continuous chromatography systems. The development of an integrated scale-down platform of perfusion culture linked to continuous capture chromatography will enable higher throughput experiments to be performed to explore the impact of critical process parameters in the continuous upstream and downstream processing steps on quality and performance.
Initially, analysis of the capabilities and limitations of existing perfusion culture mimics (e.g. with high throughput mini-bioreactors (eg ambr, Sartorius) and spin tubes) and continuous chromatography mimics (e.g. 1ml columns) will be carried out. This will form the basis of novel designs of scale-down mimics that overcome limitations with existing methodologies and have sufficient sensors integrated into the devices for e.g. pH and dissolved oxygen for cell culture. The performance of the integrated scale-down platform will be validated with bench scale (5L) perfusion runs linked to continuous chromatography at MedImmune.
Experimental data generation for integrated continuous bioprocesses
The scale down perfusion mimics will be linked to continuous capture chromatography mimics for experimental data generation linking cell culture performance with downstream performance. Key process parameters in perfusion culture (e.g. feed flowrates) and continuous chromatography (e.g. residence time, flowrates) will be identified using risk analysis. Key performance outputs (e.g. productivities) and quality attributes (e.g. glycosylation, charge isoforms, product-related impurities) to measure in each experiment will be identified. These inputs and outputs will form the basis for DoE experimentation using industrially relevant cell lines from MedImmune expressing novel modalities eg bispecific antibodies.
Chemometrics for bioprocess modelling and control of continuous processes
Experimental data will be used to build predictive cause-and-effect models that link critical process parameters (e.g. perfusion feed flowrates) to critical quality attributes (e.g. aggregates) and performance metrics (e.g. perfusion culture productivity, chromatography yield) using advanced chemometrics / multivariate data analysis techniques. The temporal impacts of continuous operation will be explored; m


10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
BB/R506205/1 01/10/2017 30/09/2021
1953111 Studentship BB/R506205/1 25/09/2017 24/09/2021 Felix Roth
Description Continuous production of biological drugs such as monoclonal antibodies (mAbs) is widely seen in industry as the next step to improve productivity and build flexible, future-proof production facilities. While there is a need for these new technologies, companies are still reluctant to adopt continuous processing due to a perceived lack of understanding of the underlying technologies, and practical difficulties leading to increased workload on operators or a higher risk of failure. In an effort to address these issues and increase the adoption of integrated continuous bioprocessing in industry, scale down models have been developed to reduce the cost and footprint of process development and allowing reliable scale-up to production scale systems.
The aim of this project is to use process data to build models that can predict process performance and the impact of operating conditions on product quality and yield. To this end, oxygen transfer data was collected from a 3 litre bench-top bioreactor and correlated to viable cell volume and viable cell density. The model was then used to estimate viable cell density in real time, a metric which can be used for a variety of process control applications such as cell density control via bleeding, perfusion rate control, or for load or productivity estimates for connected downstream operations.
This model has been implemented in a lab-scale system with the industrial collaborator and further development and applications continue.
Exploitation Route The models developed as part of this award may be used by industrial actors to improve process understanding and robustness, aid process development and reduce operator workload and failure rate in commercial processes.
Sectors Pharmaceuticals and Medical Biotechnology

Description UCL-Astrazeneca Centre of Excellence 
Organisation AstraZeneca
Country United Kingdom 
Sector Private 
PI Contribution Models of industrially relevant bioprocesses were developed to automate upstream process operation, as well as helping detect deviations and faults early. These help make AstraZeneca's upstream processes more robust and reduce operator workload.
Collaborator Contribution Astrazeneca provided access to their labs and resources, including bioreactors, use of commercial media and cell lines, and guidance and training on the equipment used. Analytical support was also provided.
Impact The collaboration has resulted in a conference presentation titled "Advanced control strategies in CHO perfusion culture for automation and optimisation of productivity and cell growth", presented at the 16th Annual bioProcessUK Conference 2019 in Liverpool, UK.
Start Year 2017