An Integrated Approach to Viral Vector Manufacturing and Supply Chain Network Development
Lead Research Organisation:
Imperial College London
Department Name: Chemical Engineering
Abstract
Advanced Therapy Medicinal Products (ATMPs) form a novel class of biologic therapeutics which
pave the way to prevention and treatment of life-threatening diseases. The versatility of
genetically-engineered drug delivery vehicles is also seen in preventive healthcare, with recent
approvals of vector-based vaccine platforms. Viral vectors are currently at the forefront of both
next-generation vaccines and therapeutics; the unprecedented demand for viral vectors forces
manufacturers to simultaneously tackle engineering product and process-related challenges, while
scaling up their production. This is highlighting the need for sophisticated decision-making tools,
which enable effective manufacturing and distribution planning throughout product lifetimes.
Process systems engineering (PSE) has traditionally assisted the pharmaceutical, and more
generally, process industries in the development of such tools. Despite the absence of works
addressing the novel viral vector industry, a review of literature to date reveals classes of
transferable PSE tools for planning of single- and multi-site manufacturing and distribution networks;
these often address complexities of flexible multi-product and multi-purpose manufacture.
Computational challenges emerge alongside the problem of planning under uncertainty, which is
considered in frameworks for capacity planning and portfolio management for the pharmaceutical
industry. Limitations in well-established approaches emerge and the opportunity brought by
application-driven tools for the viral vector industry to operations research is highlighted.
A series of research aims and objectives is proposed, followed by an overview of the progress
to date in light of said objectives. A preliminary mixed-integer linear programming (MILP) model
for the design and optimisation of UK viral vector supply chains is presented and results illustrate
the capabilities and areas of improvement of the developed snapshot model. The framework in
turn relies on techno-economic analyses obtained from computer-aided modelling of well-established
viral vector manufacturing protocols. A local sensitivity analysis highlights a need to systematically
account for uncertainty of the assumed inputs when characterising model outputs.
pave the way to prevention and treatment of life-threatening diseases. The versatility of
genetically-engineered drug delivery vehicles is also seen in preventive healthcare, with recent
approvals of vector-based vaccine platforms. Viral vectors are currently at the forefront of both
next-generation vaccines and therapeutics; the unprecedented demand for viral vectors forces
manufacturers to simultaneously tackle engineering product and process-related challenges, while
scaling up their production. This is highlighting the need for sophisticated decision-making tools,
which enable effective manufacturing and distribution planning throughout product lifetimes.
Process systems engineering (PSE) has traditionally assisted the pharmaceutical, and more
generally, process industries in the development of such tools. Despite the absence of works
addressing the novel viral vector industry, a review of literature to date reveals classes of
transferable PSE tools for planning of single- and multi-site manufacturing and distribution networks;
these often address complexities of flexible multi-product and multi-purpose manufacture.
Computational challenges emerge alongside the problem of planning under uncertainty, which is
considered in frameworks for capacity planning and portfolio management for the pharmaceutical
industry. Limitations in well-established approaches emerge and the opportunity brought by
application-driven tools for the viral vector industry to operations research is highlighted.
A series of research aims and objectives is proposed, followed by an overview of the progress
to date in light of said objectives. A preliminary mixed-integer linear programming (MILP) model
for the design and optimisation of UK viral vector supply chains is presented and results illustrate
the capabilities and areas of improvement of the developed snapshot model. The framework in
turn relies on techno-economic analyses obtained from computer-aided modelling of well-established
viral vector manufacturing protocols. A local sensitivity analysis highlights a need to systematically
account for uncertainty of the assumed inputs when characterising model outputs.
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/R513052/1 | 01/10/2018 | 30/09/2023 | |||
| 2615428 | Studentship | EP/R513052/1 | 01/10/2020 | 31/03/2024 | Miriam Sarkis |
| EP/T51780X/1 | 01/10/2020 | 30/09/2025 | |||
| 2615428 | Studentship | EP/T51780X/1 | 01/10/2020 | 31/03/2024 | Miriam Sarkis |
| Description | Viral vectors are advanced therapy products used as genetic information carriers in vaccine and cell therapy development and manufacturing. Viral vector manufacturing has still not reached the level of maturity of biologics and is still highly susceptible to process uncertainties such as viral titers and chromatography yields which challenge capacity planning exercises. The COVID-19 pandemic has further challenged viral vector manufacturers as they were pressured to respond to global demand of vaccines in a timely manner, repurpose existing capacity. This adversely impacted cell and gene therapy manufacturing, with resulting shortages and delays in viral vector raw material supplies. This highlighted the lack of a systematic framework and approach to support capacity planning under process-related and demand uncertainty within this emerging sector. In my PhD I am developing a computer-aided tool to support decision-making in this space and identify resilient planning strategies that mitigate risks of failing to meet demands and loss of investment. To do this I developed (i) a methodology quantification and characterisation of scale-dependent process uncertainties, (ii) an optimisation-based decision support tool which explores alternatives and trade-offs that emerge during scale up and identifies the required investment, production and distribution plans to fulfill demands with minimised costs. Specific milestones related to my research are listed below: • I have developed techno-economic models to evaluate and compare industrially relevant vector platforms. This enabled the quantification of key process performance indicators for different vector applications (vaccine, gene therapy) and highlighted underlying differences in throughputs and unit production costs. Results were found to be in agreement with published data for different vectors and scales of manufacturing. • I have implemented uncertainty and global sensitivity analyses on each vector platform process flowsheet to quantify and characterise the impact of underlying manufacturing uncertainties on Key Performance Indicators (KPIs) of interest, which are expected to drive investment decisions. This serves as a tool to identify pressure points in manufacturing platforms and the scalability of manufacturing assets. Additionally, through the identified uncertainty ranges for throughput and cost performance at different scales, the constructed data sets can be integrated in the development of pro-active manufacturing strategies which minimise risk of failure to meet demand and lose investments. • I have developed a supply chain planning optimisation-based framework which explore the manufacturing strategies for viral vectors, namely alternative investments, scales selected and scheduling plans. Good candidate plans are found through the selection of a suitable performance metric, which is cost and/or product availability. Given information on scale-dependent costs, logistics costs, a series of locations, scalability constraints and a set of products demands to fulfil, the optimisation identifies capacity plans, production schedules which minimise costs and meet product demands. Based on a range of case studies, viral vector manufacturing is driven by economies of scale, whether vectors are manufactured for vaccine applications or personalised therapeutics. The next step until completion of the award is to systematically integrate process related and demand uncertainty within the optimisation routine. This effectively translates in accounting for a range of process performance at different scales and a range of demand scenarios and rely on optimisation principles to identify good candidate supply chain strategies that are successful at meeting product demands with minimised risks of failure and costs. |
| Exploitation Route | The developed tool is aimed at supporting the development of agile and resilient supply chains for viral vectors. Viral vectors are currently at the forefront of ATMP manufacturing and delivery, however the methodologies developed in my work are transferable to other product applications involving therapeutics under development with future successful applications in a range of markets (from vaccine, to target therapeutics to personalised medicine). These may range from mRNA-based, pDNA-based therapeutics or novel non-viral gene delivery methods. Given the promising clinical outcomes of gene therapy, demand for these products will continue to grow rapidly. These emerging industries may face similar challenges related to scaling-up production with demand fluctuations, clinical outcomes uncertainty and uncertainty around future technology capabilities and would experience the need to identify resilient strategies forward to make groundbreaking therapies available to those in need, with minimised risks. Additionally, besides the specific application, my contribution with respect to integrating process uncertainty systematically in the optimization routine for supply chain planning will be taken forward by other researchers within the operations research field. The concept of supply chain resilience is a topical one, with current research focusing on developing decision-support tools of suitable modelling complexity, tractable computational effort and considering supply chain uncertainties including manufacturing uncertainties, demand and process disruptions. |
| Sectors | Healthcare,Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology |
| Title | Supply chain optimisation model |
| Description | The supply chain optimisation model for viral vector supply chain is aimed at assisting decision makers in finding cost-effective and successful planning strategies to meet location dependent demands. It is developed in Python and is based on principles of mixed integer linear programming (MILP). The optimisation framework explores a range of alternative decisions including (i) investment-related decisions - productions scale selection, supply chain connections and links (ii) operational decisions - sequencing of tasks to meet production targets and fulfill forecasted demands. The aim of the optimisation is to identify good candidate supply chain structures, select suitable production scales and schedule operations to meet demands and minimise costs. Additional supply chain performance indicators can be integrated, which may quantify environmental footprints of candidate solutions or resilience. The optimisation problem leverages on sets of input data. The required dataset includes techno-economic performance (process throughput and costs), process times and batch sizes at different scales for a range of products. This knowledge is constructed for a range of viral-vector based products via the techno-economic modelling exercise. Additional inputs related to candidate manufacturing locations, storage locations and demand zones (e.g. hospitals) are sourced from literature and industrial reports. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | I have presented this model at the 32nd European Symposium for Computer Aided Process Engineering. This has steered additional discussions on how to model operations throughout the supply chain and how to integrated process uncertainty in systematic fashion, so as to identify optimal supply chain planning decisions which hedge against this uncertainty. |
| URL | https://www.scopus.com/record/display.uri?eid=2-s2.0-85135395801&origin=inward&txGid=941dc405477d9f0... |
| Title | Techno-economic model for viral vector manufacturing applications |
| Description | I have developed techno-economic process flowsheet for viral vector applications at different manufacturing scales. This allows quantification and comparisons of process performance in terms of throughputs and costs. The models are developed in SuperPro Designer and are integrated with a MATLAB - SobolGSA interface which enables uncertainty analysis and global sensitivity analysis implementations. Through this computational setup input uncertainty can be systematically considered and the impact on techno-economic outputs can be quantified and characterised. Specifically, probability distributions for each of the inputs can be specified and the routine generates input samples which are simulated iteratively to retrieve output samples. This enables the quantification of uncertainty ranges for each output metric (e.g. process costs, throughputs) and computation of sensitivity indices which indicate the significance of each input variability on the observed output variability. The models are integrated within a framework which samples input uncertainty and can be used to retrieve outputs and quantify input-output sensitivity. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2022 |
| Provided To Others? | No |
| Impact | The work has enables the quantification of scale-dependent uncertainty and scalability of manufacturing assets for a number of products. This can assist investment planning decisions and de-risking scale up strategies. The models and framework area included in the submitted publication, where the use-case of the framework is extensively discussed. Characterization of key manufacturing uncertainties in next generation therapeutics and vaccines across scales. M. Sarkis, N. Shah, Maria M. Papathanasiou, Submitted Additionally, the process-related uncertainty dataset is to be used within the optimisation framework for carrying out supply chain optimisation under process-related uncertainty and identify supply chain solutions with improved resilience. |
| Description | Process uncertainty characterization for viral and non-viral cell and gene therapy products |
| Organisation | Imperial College London |
| Department | Centre for Process Systems Engineering |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | Provided techno-economic modelling framework, uncertainty analysis and sensitivity analysis results for the viral-based pathway of cell and gene therapy products. Collaborated on writing a journal paper publication comparing non-viral and viral vector pathways and techno-economic performance under uncertainty. Additionally, we have collaborated in the writing of a book chapter on integrated process design and optimization, leveraging on the acquired in-depth understanding of the impact of underlying manufacturing uncertainties to process performance. |
| Collaborator Contribution | Provided modelling framework, uncertainty analysis and sensitivity analysis results for the non viral-based pathway of cell and gene therapy products. Collaborated on writing a journal paper publication comparing non-viral and viral vector pathways and techno-economic performance under uncertainty. Additionally, we have collaborated in the writing of a book chapter on integrated process design and optimization, leveraging on the acquired in-depth understanding of the impact of underlying manufacturing uncertainties to process performance. |
| Impact | Journal Paper: Uncertainty quantification for gene delivery methods in cell and gene therapy: a manufacturing roadmap from Phase I clinical trials to commercialization. N. Triantafyllou, M. Sarkis, A. Krassakopoulou, N. Shah, Maria M. Papathanasiou, C. Kontoravdi. Submitted Book Chapter: Integrated process and supply chain design and optimization N. Triantafyllou, M. Sarkis, N. Shah, C. Kontoravdi, Maria M. Papathanasiou. Submitted |
| Start Year | 2022 |
| Description | Participation in Future Targeted Healthcare Hub Meetings |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | I presented my work at the Future Targeted Healthcare Manufacturing Hub Specialist Working Groups meeting on 31/10/2022. The meeting focused on researchers sharing cost modelling, supply chain analysis and digital twins for CAR T-cell therapies, viral vectors and mRNA platforms. Additionally, perspectives on sustainability and net zero modelling were discussed. I presented my work on integrating viral vector supply chain planning with CART cell therapy manufacturing and insights on integrating sustainability in capacity planning, which is work in collaboration with on-going undergraduate and postgraduate students whom I am supervising throughout the PhD programme. This steered fruitful roundtable discussions in the second part of the meeting. I have also attended previous meetings and found opportunities to expand my network and meet representatives from industry. This has represented an invaluable opportunity to tailor aims and objectives of my PhD to industrially-relevant open challenges. |
| Year(s) Of Engagement Activity | 2021,2022,2023 |
| URL | https://www.ucl.ac.uk/biochemical-engineering/research/research-and-training-centres/future-targeted... |
| Description | Podcast Interview |
| Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Postgraduate students |
| Results and Impact | I participated in one of the episodes of the Imperial ChemEng Cast: Researchers in focus. The podcast interviews researchers from around the department of Chemical Engineering to find out more about their work. I spoke about my journey to Imperial, my current work on pharmaceutical supply chain optimisation and I explain to a non-scientific audience the significance of my research. |
| Year(s) Of Engagement Activity | 2022,2023 |
| URL | https://lnkd.in/edmZi3Yz |
| Description | Sargent Centre for Process Systems Engineering Industrial Consortium |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Industry/Business |
| Results and Impact | I presented my work as a poster in the 2022 Industrial Consortium for the Sargent Centre for Process Systems Engineering. PhD students from the research centre participating in the poster session were to prepare a 3 minute presentation to pitch their research to Industrial partners joining the meeting and continue discussion at their poster. It was an important opportunity to gain feedback from academics within operations research and process systems engineering and steer discussions towards to common challenges encountered in modelling and optimisation of complex supply chains. |
| Year(s) Of Engagement Activity | 2022 |
| URL | https://www.imperial.ac.uk/process-systems-engineering/industrial-consortium/annual-industrial-conso... |
| Description | Sargent Centre for Process Systems Engineering Student Committee |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Postgraduate students |
| Results and Impact | My roles as Communications Officer (21-22) and Chair (22-23) in the Sargent Centre Student Committee have contributed to the organization of social and networking events and student-led academic seminars which restored a sense of community within the multi-institutional (Imperial College - UCL) research centre post-pandemic. |
| Year(s) Of Engagement Activity | 2021,2022,2023 |
| URL | https://twitter.com/sargent_centre |