EPSRC Fellowships in Manufacturing - Macromolecular Manufacturing

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

Abstract

The bioprocess industry manufactures novel macromolecular drugs, proteins, to address a broad range of chronic and debilitating human diseases. The UK holds a leading position by virtue of its science base and has unique university capabilities underpinning the sector. Whilst revenues are large, ~£110bn in 2009 on a worldwide basis, there are huge pressures on the industry for change if demands for cost reduction and waste minimisation are to be met, and populations are to benefit from the potent drugs becoming available. A sea change in manufacturing will be needed over the next decade or so if the potential of modern drugs are to make their way through to widespread distribution. It has been estimated that an economic return of 200-1000% can be gained by applying modelling and optimisation tools to bioprocessing.

In this fellowship I will address a series of related issues which will be necessary in order to achieve the transformation in manufacturing necessary for the UK to build a world lead. Broadly, there are two goals:-

i) to create mathematical models of complete manufacturing processes used for macromolecule production; and
ii) to use these mathematical descriptions as tools for the faster prediction of efficient and reliable operation so that industry can be more flexible in the way it runs intrinsically costly process plants.

Research to address this first goal will require collaborations with a range of scientists, mathematical, engineers and computer scientists in order to work out the most effective way to describe complex processes in mathematical equations. Of particular importance will be the need to devise methods which allow us to capture how the process for making a particular material actually changes the material properties. This "process memory" is a poorly understood phenomena but has a profound impact on manufacturing performance.

The second goal will be addressed via an industrial forum providing materials and data for testing of our work. The most crucial aspect will be the need to use the mathematical description of the process as a framework for setting up control strategies designed to be reliable even if the manufacturing process is perturbed. Such events are frequent in biological systems where inherent levels of variability mean that the processes are never truly identical.

The project, as a whole, will therefore lead to the rational development of robust processes that should work with minimal change during manufacturing. This, in turn, will lead to reductions in manufacturing costs; such reductions are critical to producing the next generation of protein and vaccine medicines in quantities sufficient to meet worldwide demand.

Planned Impact

The benefits of the proposed EPSRC Manufacturing Fellowships will be significant. This is a radical new opportunity for the industry which suffers from a lack of joined up thinking and tends to operate in discrete silos of expertise whereas an integrated approach would pay high dividends.

UK-based companies will benefit from access to a research activities created by the fellowship. The research will allow UK companies to understand better the likely routes for manufacturing of their medicines and to have, for the first time, a genuine capacity to achieve robust control in spite of biological variability. Manufacturing efficiencies will be raised and waste reduced. The fellowship will create a network to provide a conduit for effective knowledge exchange. A key metric of success will be application in industrial practice and the adoption of the methods created. Industry will also gain from access to the skilled researchers who will form an integral part of the fellowship and who will each have benefited from a wide and interdisciplinary approach to their research and its application.

Potential patients will benefit because the fellowship innovations will significantly aid reduction in development times of macromolecular medicines, which is particularly crucial for those addressing previously unmet clinical needs and the treatment of severe conditions such as arthritis and cardiovascular disease, as well as vaccines for previously unpreventable viral infections and cancers. By providing industry the capabilities and tools to achieve changes to manufacturing processes we shall, for the first time, open up possibilities for major improvements to processes during production and hence reduce costs to the NHS. The capacity to treat conditions such as rheumatoid arthritis much more effectively in ageing populations is vital, but still poses a problem with respect to stretched NHS budgets. A significantly greater number of drugs will be capable of meeting NICE's thresholds and thus benefit extended patient populations.

The UK economy will benefit because academic research will complement the country's strength in bioscience discovery. Collaboration between bioprocess engineers, process modellers manufacturing experts, regulators and physical scientists will ensure effective knowledge and skills transfer between the science and engineering base and UK industry and the regulating agencies. This will expand the UK position in the global healthcare market and attract further R&D investment from global business which recognises the UK as a good place to conduct these activities.

Macromolecular medicines are complex and labile so that bioprocess development times and costs tend to be high due to unforeseen issues that occur during scale-up of the manufacturing processes. Currently there is little scope to alter a manufacturing processes because we cannot readily predict the effects that such changes will have. This transformative research agenda will allow for the first time engineers to control directly the quality of output during manufacture. By creating and then testing manufacturing models for whole bioprocesses we shall gain fundamental engineering insights crucial for the more effective direction of acquisitions of experimental data and also the improved design and operation of whole bioprocesses. These will together raise manufacturing efficiencies and reduce waste. Such a vision is consistent with recent efforts by the regulatory authorities, and in particular the Quality by Design (QbD) initiative of the International Committee on Harmonisation (ICH), to develop science-based regulatory submissions for approval to manufacture new biological products. The research will allow UK companies to gain a better prediction of the likely routes for manufacture of medicines. The tools will enable the UK biotechnology industry to design more efficient bioprocesses that fit to available as well as emerging manufacturing.

Publications

10 25 50

publication icon
Fischer Vivien (2018) Optimisation of preparative polishing chromatography for a ternary antibody mixture in ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY

publication icon
Goh Hai-Yuan (2018) Rapid process monitoring & control in mammalian cell culture using off-gas mass spectrometry analysis in ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY

 
Description We have made improved methods to optimise the efficiency of purification of therapeutics. These methods have been used in collaborations with international biotechnology organisations. We are continuing to improve the optimisation method, especially on the identification of effective, near-optimal operating conditions in situations where there are multiple optima. Such situations are not uncommon in engineering problems, especially those with a fair amount of noise. We have so far been able to find conditions that are within 90% of the global optimum in all cases (and have actually identified the global optimum in more than 80% of the cases tested so far). We have also shown that we can search simultaneously across continuous and discrete variables; further study is ongoing to see if these novel approaches can be generalised. If so, these approaches will be extremely interesting to many research and development groups, in biotechnology and healthcare and possibly beyond. We have recently completed a novel and complete analysis of continuous countercurrent chromatography, part of which was just accepted for publication. This analysis will allow researchers, for the first time, to arrive rapidly at a near-optimal set of operating conditions for protein purification in this method.
As summarised elsewhere, our whole-bioprocess work on the production of a monoclonal antibody--from its production in a 50-litre bioreactor, through recovery, purification, and final freeze-drying--has led to valuable results that are currently being prepared for publication. This paper will provide a valuable instance of the interactions that arise among steps in a complex bioprocess, and how these interactions can be dealt with to maintain the quality and efficiency of the process.
Exploitation Route We cannot describe this fully, because the project is still on-going. But an important aspect of our approach is to identify efficient, near-optimal operating conditions for complex bioprocess steps in a way that is more rapid and effective than the conventional experimental designs. Once we have demonstrated our approach for a wide range of industrially relevant datasets, it is likely that the method would be of great value to biopharmaceutical process development, and therefore is likely to be taken up by biotechnology, biopharmaceutical, and pharmaceutical companies.
Sectors Healthcare,Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology

 
Description Several biotechnology companies are interested in the results my group is producing. This has led to collaborations and consultations. We are working with Merck and Co, New Jersey, USA, on the use of empirical optimisation models (refinements of the adaptive simplex process) for the rapid and effective development of chromatographic unit-operations. I am also consulting for Bristol-Myers Squibb, New York, USA on the use of mechanistic models for the optimisation of polishing chromatography steps. Discussions are on-going with other organisations, including GE Healthcare, Inc. The immediate impact of this work is in improving the workflow of bioprocess development in industry: both in facilitating the development process (shortening the duration and allowing for high-throughput implementation) and in arriving at more robust and effective end-points. As mentioned in the summary, some of the final results from the research of this Fellowship are being analysed now, and should be submitted for publication shortly. These results should enhance collaboration and uptake by the bioprocessing industry. Examples of such uptake include the hiring of a post-doctoral associate by Merck and Co, a company with whom we had published two papers on the use of the adaptive simplex method (papers already reported). Further, the adaptive simplex method has been used by an engineering doctorate student who worked in collaboration with GSK, Stevenage, UK; the associated publication has been reported by the student in a separate project, to which this document has been linked. In this case, again, the student was hired by GSK after the completion of the EngD.
First Year Of Impact 2017
Sector Healthcare,Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology
Impact Types Societal,Economic

 
Description EPSRC general call
Amount £2,961,745 (GBP)
Funding ID EP/N024796/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 07/2017 
End 08/2020
 
Description Open EPSRC application.
Amount £2,961,745 (GBP)
Funding ID EP/N024796/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 09/2016 
End 08/2021
 
Title New method of experimental multivariate optimisation 
Description An optimisation method called the adaptive simplex method was developed as a numerical tool for optimisation in the 1960s. Its application to experimental, as opposed to numerical, optimisation was realised in the 1970s, and applications to a variety of scientific areas, especially in the area of chemometrics, began then. However, its application to multi-objective problems has been very limited, with significant results only having been achieved in the last 15 years or so. Our approach has been quite different from those in the published literature. Our approach has so far been extremely successful not only in finding effective operating conditions, but in finding the global optimum in multiple-optima problems very often. It is now being tested on many experimental systems, including industrially relevant datasets from our industrial collaborators. 
Type Of Material Improvements to research infrastructure 
Year Produced 2016 
Provided To Others? Yes  
Impact An outline of our innovations in this area was published in the article "Application of simplex-based experimental optimization to challenging bioprocess development problems: Case studies in downstream processing", by Konstantinidis et al (2016).The general method, as applied to downstream processing problems, was published in the article "Data-driven multi-objective optimization via grid compatible simplex technique and desirability approach for challenging high throughput chromatography applications" by Konstantinidis et al, Biotechnology Progress (2018). Both articles have been included in the list of relevant publications. Once the general applicability of the method has been developed, it is likely to be readily used in a variety of problems. Ideally, it will become an alternative to studies based on experimental design in the early development of processes (our focus has been on bioprocesses, but the method is quite general). In particular, it may be useful in healthcare studies, and in that context might reduce the number of experiments needed in animal studies. The method has since been used in other research studies. Vivien Fischer, a doctoral student at UCL who worked on an Engineering Doctorate, used the simplex method among other tools in developing a comprehensive optimisation method for the two-step polishing chromatography of proteins. A link has been provided to Fischer's project. 
 
Description Novel development methods for bioprocess development 
Organisation Merck
Country Germany 
Sector Private 
PI Contribution I have been working with the bioprocess development team at Merck, New Jersey, USA to develop new methods to arrive rapidly at robust and effective operating conditions for several bioprocess steps, including chromatography, protein folding, and cell culture. Of particular interest is the use of our version of the adaptive simplex method of empirical multivariate optimisation for the rapid identification of effective, near-optimal operating conditions for a range of industrially relevant experimental datasets. These datasets, all of which are on complex proteins, including polishing chromatography, refolding of proteins from inclusion bodies, and cell culture.
Collaborator Contribution Merck has provided datasets on several protein-based materials (either marketed products or candidates in the process towards licensure). These include data on polishing chromatography, refolding of proteins from inclusion bodies, and cell culture.
Impact We have published one paper on the use of our version of the adaptive simplex method applied to several datasets, all of which are on complex proteins, including polishing chromatography, refolding of proteins from inclusion bodies, and cell culture. This paper was published in Biotechnology Progress in 2016 (full details under the publication list).
Start Year 2014
 
Description Preparative chromatography of complex biologics 
Organisation Bristol-Myers Squibb
Country United States 
Sector Private 
PI Contribution Bristol-Myers Squibb (BMS) shared an experimental dataset with me to analyse the interactions of closely related biologics in the purification of complex medicines. Analyses on this dataset are ongoing, but insights have been obtained that should be immediately relevant to many other problems in the purification of next-generation medicines. Once the results have been approved by the BMS legal team, this should result in an important publication.
Collaborator Contribution As mentioned above, BMS shared a proprietary dataset of a kind that is not easy to generate in academia.
Impact The expected impact, once the research article has been published, is relevant to the efficient production of complex medicines, and is therefore expected to have an economic impact.
Start Year 2017
 
Description Biotechnology conference (India) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact I gave a keynote presentation on the scope of biological separations in the healthcare industry. The audience, which included biotechnology professionals from South-East Asia, faculty from around the world (including the USA and Germany), and undergraduate and post-graduate students at the Indian Institute of Technology Madras, in Chennai, India, were interested and asked many questions. I am now discussing possible collaborations with some of the industrial organisations who were represented at the meeting.
Year(s) Of Engagement Activity 2010,2015
 
Description Conference (Salzburg, Vienna, 2016) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact I presented research work from my group on novel empirical optimisation methods in biotechnology at an internationally recognised conference. Our approach was found interesting and novel by many industrial practitioners, and some companies discussed potential methods to adopt it in their design methodology.
Year(s) Of Engagement Activity 2016
 
Description Recovery Conference 2016 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact I presented research findings in the presentation entitled "Refined simplex as superior process development method for bioprocess steps and sequences". It was widely appreciated at this select conference at which the major industrial and academic contributors to biological downstream processing were present.
Year(s) Of Engagement Activity 2016