Section 1: Sean Chang discusses Thermo Fisher’s solutions for CAR T manufacturing & introduces DeltaV

Considerations & benefits of a closed, modular cell therapy workflow  

The key issues in cell therapy manufacturing today can be illustrated within the field of autologous T cell therapy. The manufacturing process is complex, labor-intensive, and requires many open manipulations. It is also difficult to synchronize different instruments and products to make the workflow traceable and compliant with regulatory requirements.

To solve these issues, we propose three main solutions. Firstly, a closed system will minimize contamination. Secondly, a modular system will maintain flexibility. Third and most importantly, automating the process will reduce labor and human error. It is in this third aspect that digitalization plays a particularly vital role.

To support cell therapy manufacturing, Thermo Fisher Scientific is building closed, modular and automated instruments, which can be incorporated into a cell processing platform and controlled by the DeltaV™ distributed control system (Figure 1Instrument solutions for every step of the T cell therapy workflow.). Digital integration will allow fully automated process management and maintain data integrity. Consequently, delivering a software solution that provides direct connection to the DeltaV is a key step in the platform’s ongoing development.

Thermo Fisher solutions for CAR T manufacturing

Instrumentation for each part of the workflow has been developed to address challenges across the entire end-to-end process, beginning with cell processing (wash, concentrate, buffer exchange) using the CTS Rotea™ system followed by isolation and activation using the Invitrogen Dynabeads magnetic separation products, which include the CTS™ DynaMag™ magnet. For the cell engineering step, the newly launched CTS™ Xenon system™ for large-scale electroporation can be used, followed by cell expansion, employing a high-throughput mode rocker and controller for dynamic culture. At the end of the process, cells can be harvested and processed for fill and finish using the Rotea, and cryopreservation can be performed using the CryoMed^® control rate freezer. Throughout the manufacturing processes, incubators designed and certified for cleanroom use for static culture are available for use. All of these instruments are being adapted and designed to have the capability of connecting to DeltaV systems directly, allowing them to be controlled and managed in the same network within the same interface.

Proof of principle studies performed using this platform demonstrate the high efficiency of peripheral blood mononuclear cells (PBMC) and T cell isolation using the Rotea and DynaMag at the beginning of the process, and show high quality and potency of the manufactured CAR-T cells after cryopreservation.

Digital strategy: off-the-shelf connection between instruments and DeltaV controller

The end goal of a mature manufacturing environment is to manage the instrument layer through a distributed control system (DCS), enabling the integration and management of workflows, and ensuring traceable, repeatable, and secure data connectivity through manufacturing execution systems (MES) up to enterprise level (ERP). The DeltaV DCS controller is one of the most reliable and widely used DCS controllers in the biopharmaceutical industry. However, there is no current ‘off-the-shelf’ software solution between the instrument and the DCS layer. This connection requires coding, engineering, and configuration to allow the DCS layer to talk to the instruments directly. Furthermore, customized software development projects provided by tools providers or DCS vendors are usually costly and time consuming. 

Thermo Fisher Scientific’s new ‘off-the-shelf’ software product will allow customers to operate all of the modular instruments in the same DeltaV network, using the same operator interface (Figure 2The five software modules included in Thermo Fisher’s software product, providing DeltaV connectivity to Rotea™, the next-generation magnetic separation system, Xenon™, CryoMed™, and incubator modules.). Utilizing DeltaV’s capabilities, this software product allows the user to create batch recipes to control different workflows across various unit operations, and to collect and store historical data in compliance with regulations. The user will also be able to retrieve data to produce batch reports through third-party batch reporting packages.

The software includes an OPC-unified architecture (UA) interface module, an equipment module, and phases (the building blocks of batch recipes) which run in the DeltaV controller. The interface module maps the data between DeltaV and the instruments, whilst the equipment module executes commands to the instruments. The equipment module can be controlled by higher level batch recipes using phases. The DeltaV Batch Executive is used to create batch recipes for different workflows and collect data from all instruments. The software interface includes batch banners that can be used for messages and prompts from both equipment modules and workflow phases, and an equipment module faceplate that can be used for manual control – for example, to start or stop a protocol.

Section 2: Bruce Greenwald disscusses DeltaV architecture & system start-up costs

Building your digital plant with DeltaV

The BioPhorum Operations Group has created the Digital Plant Maturity Model (Figure 3The digital plant maturity model.) to define the stages of maturity from paper-based plants to fully automated and adaptive plants. DeltaV and the Emerson product line can support customers all the way from the pre-digital plant level to the adaptive plant level, without replacing system components.

System start-up costs: integrating islands of automation

There is a hidden cost of integrating at an individual level. Using individual unit operations to piece together a solution to reach the historical data level requires engineering and validation at each touchpoint. This can be very costly and can severely impact the time to market. Often, layers get skipped, causing intermediate gaps that lead to paper-on-glass solutions that can impact time and cost, and make the actual day-to-day operations more complex.

DeltaV architecture

Since its release in the late 1990s, DeltaV has established the concept of ‘Easy’, due to its inclusion of ‘off-the-shelf’-type technologies. It is designed to be easy to use for the engineers who develop the system, as well as for the operators to control and maintain the system, and it is easy to get the information out of the system. DeltaV was deployed on Microsoft operating systems and off-the-shelf PCs, because in the long-term, being able to use the ‘off-the-shelf’ technologies’ embedded functionalities would make things easier for automation engineers and production personnel.
DeltaV has embedded advance process control and a built-in ISA-S88 batch process control infrastructure.

The overall DeltaV architecture is easy, flat, and simple (Figure 4DeltaV architecture.). Peer-to-peer architecture with DeltaV is used, which is not dependent on client server architecture. This allows embedded nodes to publish information, which can either be consumed by other embedded nodes or consumed at the workstation level, making system configuration easy. DeltaV provides a single, integrated, automated solution for immuno-oncology subsystems, controllers, user management, operations experience, advanced process control, and recipe management. DeltaV’s use of a single database architecture means you only have to go to a single application when additions or modifications to your system are required.

Data contextualization at the runtime level is a large part of the DeltaV solution. The system allows real-time alarm, event, and batch data to be contextualized, making it easy to share and store data in standardized databases and use industry standard tools to move data to higher levels within the system from an MES and ERP perspective.

DeltaV’s integrated capability for meeting electronic records management (ERM) and data integrity requirements in process automation applications is described in a white paper published in April 2017 [1]DeltaV™ Capabilities for Electronic Records Management and Data Integrity. Emerson. 2017. . This paper explains the configuration requirements, the real-time runtime environment, and the historization requirements, and how DeltaV complies to both US FDA 21 CFR Part 11 and EU Annex 11.

OPC-UA enables digital transformation and allows embedded nodes to function as OPC-UA servers that feed their data directly into DeltaV. DeltaV can then feed that information to third-party historians, cloud applications, reliability applications, and other analytical tools to evaluate the data that has been collected and harmonized at the DeltaV level.

Within the overall workflow, the DeltaV system sits at Level 2 (Figure 5Purdue model for process suites to business systems integration.). At Level 3, there is the laboratory information management system (LIMS), and an MES such as Emerson’s Syncade. At level 4, there is the business network where the ERP systems reside, and advanced analytics and scheduling can be performed and passed down in real-time to the DeltaV system.

Section 3: Krish Roy discusses the quality control in automation in cell therapy manufacturing, & cell manufacturing of the future

Quality-by-design-driven scalable manufacturing of therapeutic cells

From the discovery-centric perspective, cell therapy is thought of as an interaction between multiscale dynamic complex systems. The starting material is highly variable and highly dependent on prior therapies, as well as the age, sex, lifestyle, and environment of the patient (and the donor, in the case of allogeneic cell therapy). The material is used in a complex manufacturing process in which any manipulation impacts the properties and potentially the function of the cells. The product is hugely complex compared to anything the biopharmaceutical industry has ever manufactured before. Furthermore, once the engineered cells are delivered to the patient, the patient’s own microenvironment shapes their properties, behavior, and function.

The complexity from the basic donor side to the patient side poses a tremendous data challenge. How to create models to predict whether particular cells are going to be functional in a specific patient with a specific disease remains a mystery. This is where digitization, data integration, and data processing are of tremendous value.

There are two key areas where large data processing and integration, and thus digitization, are needed. Firstly, in the identification of critical process parameters (CPPs) and monitoring of early quality attributes of the manufacturing process to ensure a consistent, reproducible product within specific parameters. Secondly, in the identification of the multivariate parameters of a product that are the most predictive of patient outcome. Both require an understanding of data manipulation, analytics, data sciences, and data collection for very large data sets.

In-line or at-line process & product analytical technologies

In- and at-line process analytical measurement testing and product quality control (QC) will be the future of the cell manufacturing industry. Right now, processes are very fixed and recipes are repeated, despite the starting material being so varied. (Even for allogeneic therapies, the starting material will differ from donor to donor).

In-line or at-line process analytical technologies during R&D and process development allow discovery of early product critical quality attributes (CQAs) to predict end-product quality, and discovery of CPPs that control end-product CQAs. During manufacturing, process analytical technology (PAT) allows the early identification of batch failure, monitoring of microbial contamination, and monitoring of CQAs and CPPs to ensure optimal CMC compliance. Feedback controls, data management, data integration, and digitization become critical.

Cell manufacturing of the future: product is the product

In this vision of the next generation of cell manufacturing, the product is the product, rather than process being the product. Within the next decade, bioreactors should have multiple sensors and measurement tools, and digitalization and data input will become critical in supply chain management and logistics.

The field will see digital models of both a centralized and a distributed cell manufacturing network, and capacity planning tools that select the optimal locations and manufacturing capacities. Impacts of reagent supply disruptions and labor shortages on patient access and capacity utilization must be considered. A hybrid cost model including activity-based costing and parametric costing will be needed. This will be used to assess cost implications and return-on-investment of technological innovations. Data-driven manufacturing, PAT, supply chain management, artificial intelligence (AI), machine learning (ML), and automation will all be critical elements to be integrated within the digitalization and digital infrastructure process.

Data infrastructure needs to be layered into both the physical and the human infrastructure, alongside the need for an integrated conduit between the three that includes sensor-controlled automation, a collaborative environment between industry and academia, preclinical data, pilot manufacturing, predictive analytics, and supply chain understanding.

References

1.DeltaV™ Capabilities for Electronic Records Management and Data Integrity. Emerson. 2017.  Crossref

Ask the experts 

Elisa Manzotti speaks to (from left to right) Sean Chang, Bruce Greenwald and Krish Roy, who answer your questions about how a fully automated cell therapy manufacturing process that addresses digital connectivity and instrument-to-instrument compatibility can increase quality of the final product and reduce
manufacturing failure rates.  

How does DeltaV provide data integrity for configuration data?  

BG: DeltaV contains a tool known as version control audit trail (VCAT). VCAT allows us to version every object in the configuration database. Every time someone needs to make a change, that object is checked out, the change is made, the object is checked back in, and a new version is automatically created. From a management perspective, you can see each of the different versions and who made the change. You can also compare the versions both visually and textually automatically using the VCAT tool for validation and management of change requirements. 

How would digitization and end-to-end data integration help accelerate development of cell and gene therapy (CGT) products? 

KR: It would help in multiple ways. In the scientific lens, it starts at the discovery and process development stage, where you can look at large datasets and create decision processes about quality attributes and clinical process parameters. This leads to interfacing with clinical trials and understanding mechanisms of action and the critical quality attributes that are predictive of patient outcomes. Integrating process analytics, supply chain data, and cost modelling components would greatly improve the process and the product quality and reproducibility, reduce batch failures, and drive down cost.  

I remember listening to a talk by the former president of Intel, whose processes are incredible. They have been making microchips for decades now. They have a process model, and data for every manufacturing run is fed back into the process model to further improve it. We do not do that in the biopharmaceutical industry anywhere. We need to make that move towards full digitization.  

Can Thermo Fisher’s instruments connect to other non-DeltaV DCS systems?  

SC: Yes. At Thermo Fisher, we make sure all our cell therapy instruments will be equipped with the OPC-UA which is the standard interface to allow this instrument to exchange data with other platforms or control systems. With OPC-UA, the Thermo Fisher instrument has the capability to connect to other systems. 

However, the ‘off-the-shelf’ product I introduced only provides the codes to directly connect to DeltaV. If the customer wants to connect to other DCS systems from other vendors, there is some software engineering that the customer will need to figure out with those vendors. Themo Fisher can provide support by providing the OPC-UA document manual or related information from the instrument support teams. 

With the connections to other levels within my organization, how does DeltaV manage cyber security?

BG: DeltaV can provide a bubble around your entire control system. We are compliant and allow end users to achieve ISASecure SSA Level 1 certification for their control system from a cyber security perspective. The up and out communications go through our secure Emerson smart firewall. Using industry standards like OPC-UA, we also have web services tools that will be used for connecting to the ERP or MES layers.   

What areas do developers and vendors working on digitization need to pay attention to?  

KR: When I see folks working on digitization, a lot of focus is on the manufacturing and GMP end. That is great, but we need to bring this concept down to the discovery and product development sides. Digitization needs to start much earlier if we are to understand CQAs and how product behaves under different manufacturing scales.  

We are an academic consortium of eight or nine universities, and one of the things we are trying to do is create digitization in each laboratory. This feeds into a cloud platform and allows us to do analytics, bringing the power of many experiments together to make decisions and understand the data and manufacturing variabilities better. Vendors are not there yet, but that is where the biggest long-term impact in the cell and gene therapy field will be.  

By the time we have the process transferred to our manufacturing, we are too late. It takes many years for the company to then identify quality attributes, critical process parameters and MOAs. If we can bring that process up, even to the graduate student level, we will be much better off.  

Does Thermo Fisher Scientific have other 21 CFR Part 11 compliant software solutions without a DCS system?  

SC: All of our cell therapy instruments have another standalone software product to support 21 CFR Part 11, or any regulatory requirements in terms of the digital integration. We call it the Security, Audit and Electronic signature (SAE) solution. We use the same interface and functions of the existing REO software, but further support the security settings, audit trails, and e-signature functions. It is a great start-up solution for a customer who does not have a DCS system. 

Why connect to DCS and not to MES?  

SC: At Thermo Fisher Scientific, we believe in providing comprehensive digital integration to customers. We believe in the DCS layers supporting, managing, and operating instrument layers, and supporting data integrity. We want to follow the architecture of this digital integration, so right now, we are mainly focused on connecting the DCS to instrument layers.

BG: As previously mentioned, having a harmonized layer between the MES and the field instruments in the unit operations provides a common interface. It can also automate a lot of the tasks built into their implementation with DeltaV, as opposed to having MES and unelectronic workflow instructions that are only partially automated. 

Within the DeltaV, what is the feedback loop timeframe?  

BG: From a closed loop control, it is as fast as 25 milliseconds from a modular execution perspective. With the interface to unit operations at the Thermo Fisher Scientific level we were performing supervisory control, which typically would execute in the one-second timeframe. DeltaV is acting as supervisory control, passing down commands, and then the Rotea and other devices are doing the heavy lifting.

How soon will the Thermo Fisher digital platform be available for cell and gene therapy manufacturing operations?  

SC: The first module to connect Rotea to DeltaV will launch in early Q2 of 2022. We have five different modules to connect to DeltaV that will be launched at different times. However, once you purchase one, you get access to the connectivity of all five different software modules. The newer modules will be released as an update of the whole software solution.  

Do you see the field moving to in-line testing, and will technology make this possible?  

KR: We need to bring in in-line or at least at-line process testing – it is unrealistic to have all in-line, but we should have rapid at-line process testing at least. As we move into more complex products, especially induced pluripotent stem cell (iPSC)-derived products where the manufacturing timeline is four to eight weeks, you cannot just rely on end of process testing. It does not make business sense, let alone scientific sense. 

We should start with the existing technologies for pH, glucose, and lactate. But most of the sensors on the market today were developed for other purposes and do not fit our purpose for cell and gene therapy – there needs to be significant innovation in this space. I think optical sensors have a lot of potential here. Most of the work we are doing at the moment involves multiplex measurements with wireless data transmission capabilities, with those two elements combining for a process control capability.  

Especially in the autologous setting, our input – the raw materials, the cells – is different every time we manufacture something. Each time we manufacture something, if we put it in the current fixed and uncontrolled process, we will inevitably end up with a different product. In-line testing is a way to be able to understand what a product is going through, where it is going in terms of its differentiation and expansion process, and to allow tweaking to get consistent product parameters. In that sense, PAT is one of the most critical things that we can pursue as a field. 

What is the data historian used by the DeltaV platform?  

BG: There are three different types of data and data historian. For continuous data, we have our DeltaV continuous historian. For any alarms and operator events, we have a sequel based, which is an OPC-alarm and events (A&E)-based database. A third separate database is the batch historian, which is sequel-based for all batch information automatically collected within DeltaV. Those three databases comprise the historians within DeltaV, and then provide views, OPC-historical data access (HDA), and OPC-UA to remotely query information. 

Most cell and gene therapy innovation occurs in small organizations that will not have access to these tools for a decade or two. How should we effectively manage innovation and product process development in this context?  

KR: This is why I advocate heavily towards collaboration with academia and government laboratories, that are also focusing on this mission. There are a number of existing consortia. Many of our industry partners want to collaborate and extend their capabilities and reach because not everything can be done in-house, especially for small companies and businesses. We just signed a contract with NIH to become an in-depth cell characterization hub for their medicine innovation program trials. This kind of consortium partnership is critical, especially for small and medium businesses, where they do not want to build these capabilities in-house.  

Biographies

Dr Sean Chang is the Early Innovation Manager in the Cell and Gene Therapy business at Thermo Fisher Scientific and is responsible for the integration and automation of next generation cell therapy manufacturing workflow.” Prior to this role, he was a Process Development Scientist, leading manufacturing process optimization of an allogeneic CAR-T product. Sean also gained extensive experiences in new product development in his previous career role as an R&D Scientist, where he identified novel T cell genome editing and non-viral delivery solutions, and worked on new closed modular instruments. Sean received his Ph.D. in Integrative Molecular and Biomedical Sciences from Baylor College of Medicine, where he focused on the mechanisms underlying therapeutic resistance of breast cancer.

Bruce Greenwald is the DeltaV Platform Business Development Manager for Emerson Automation Solutions, located in Austin, TX. In his current role, Mr Greenwald assists customers in understanding the features and benefits of DeltaV to improve their automation experience. Mr Greenwald is a 1979 graduate of the University of Kansas with a degree in Chemical Engineering. He started his career with the Dow Chemical in Freeport, TX and joined Fisher Controls in 1983 as a systems engineer. Mr Greenwald joined the RE Mason Company, an Emerson Impact Partner, in 2000, executing PROVOX and DeltaV projects. He held several positions at RE Mason, and in 2011, rejoined Emerson. His 4 decade-long career has been focused on process control.

Dr Krishnendu (Krish) Roy received his undergraduate degree from the Indian Institute of Technology (India) followed by his MS from Boston University and his PhD in Biomedical Engineering from Johns Hopkins University. After working for 2 years at Zycos Inc., a start-up biotechnology company, Dr. Roy left his industrial position to join the Biomedical Engineering Faculty at The University of Texas at Austin in 2002, where he was most recently Professor and Fellow of the Cockrell Chair in Engineering Excellence. He left UT-Austin in July of 2013 to move to Georgia Tech. where he is the Robert A. Milton Chaired Professor in Biomedical Engineering. At Georgia Tech, he also serves as the Director of the newly established NSF Engineering Research Center (ERC) for Cell Manufacturing Technologies (CMaT) and The Marcus Center for Cell-Therapy Characterization and Manufacturing (MC3M) – as well as the Director of the Center for ImmunoEngineering. He is also the Technical Lead of the NIST/AMTech National Cell Manufacturing Consortium (NCMC), a national public-private partnership, focused on addressing the challenges and solutions for large scale manufacturing of therapeutic cells. Dr. Roy’s research interests are in the areas of scalable cell manufacturing, Immuno-engineering, stem-cell engineering and controlled drug and vaccine delivery technologies, with particular focus in biomedical materials. In recognition of his seminal contributions to these fields, Dr. Roy is elected Fellow of the American Institute for Medical and Biological Engineering (AIMBE) and the Biomedical Engineering Society (BMES). In addition, Dr. Roy has received numerous awards and honors including Young Investigator Awards from both the Controlled Release Society (CRS) and The Society for Biomaterials (SFB), NSF CAREER award, Global Indus Technovator Award from MIT, the CRS Cygnus Award etc. He is also the recipient of Best Teacher Award given by the Biomedical Engineering Students at UT-Austin and the best advisor award given by bioengineering students at Georgia Tech. He serves as a member of the Editorial Boards of the Journal of Controlled Release, the European Journal of Pharmaceutics and Biopharmaceutics, the Journal of Immunology and Regenerative Medicine, all from Elsevier, as well as the AIChE Journal of Advanced Biomanufacturing and Bioprocessing. He is a member of the Forum on Regenerative Medicine of the National Academies of Science, Engineering and Medicine (NASEM), and a Board Member of the Standards Coordinating Body (SCB) for Cell and Regenerative Therapies.

Affiliations

Sean Chang, PhD
Manager, Early Innovation, Cell and Gene Therapy, Thermo Fisher Scientific

Bruce Greenwald
DeltaV Platform Development Manager, Emerson Automation Solutions

Krishnendu Roy, PhD
Fellow, AIMBE; Fellow, BMES
Robert A. Milton Chair
Director, NSF ERC on Cell Manufacturing Technologies (CMaT)
Director, Marcus Center for Cell-Therapy Characterization and Manufacturing
Director, Center for ImmunoEngineering at Georgia Tech
Technical Lead, National Cell Manufacturing Consortium (NCMC)
The Wallace H. Coulter Dept. of Biomedical Engineering at Georgia Tech and Emory University, Georgia Institute of Technology.

Authorship & Conflict of Interest

Contributions: All named authors take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.

Acknowledgements: None.

Disclosure and potential conflicts of interest: Roy K has recieved grants or contracts from National Science Foundation; National Institutes of Health; Food and Drug Administration; Marcus Foundation; The State of Georgia; Georgia Tech Foundation. Roy K recieved consulting fees from Johnson and Johnson, LEK Consulting, Merck KGA, Terumo BCT, Clearview, Mubadala Partners, Anzu Partners, DeciBio, MIT-Singapore Cell Therapy Alliance, Johns Hopkins ImmunoEngineering Center, Standards Coordinating Body, Cord Blood Connect. Roy K recieved payment or honoraria from NSF, NIH, Cornell University, University of Minnesota, University of Washington, University of California - Davis, CHI Bioprocessing Meeting. Roy K recieved support for attending meetings and/or travel from NIH, NSF, Cord Blood Connect, Phacilitate, Biophysical Society, National Academies of Science, Engineering, and Medicine, Bioprocessing Summit, Controlled Release Society, US-Korea NSF Nanotechnology Forum, Institute of Biological Engineering. Roy K has recieved the following patents: US Patent No. 10,799,246, Awarded October 13, 2020, Methods, compositions, and devices for the occlusion of cavities and passageways, US Patent No. 10,792,044, Awarded October 6, 2020, Methods, compo. Roy K has had a leadership or fiduciary role in Alliance for Cell Therapy Now, Standards Coordinating Body, MIT-Singapore Cell Therapy Alliance, Johns Hopkins ImmunoEngineering Center. Roy K has stock or stock options in Georgia Research Alliance, State of Georgia, University of Miami, Duke University, The Sanford Foundation Hospitals, Emory University, Kansas University Medical Center, Andrews Orthopedics Florida, University of Pennsylvania, SIRpant Immuntherapeutics, In-kind contributions through the NSF ERC on Cell Manufacturing Technologies (CMaT). The authors have no other conflicts of interest.

Funding declaration: The authors received no financial support for the research, authorship and/or publication of this article.

Article & copyright information

Copyright: Published by Cell and Gene Therapy Insights under Creative Commons License Deed CC BY NC ND 4.0 which allows anyone to copy, distribute, and transmit the article provided it is properly attributed in the manner specified below. No commercial use without permission.

Attribution: Copyright © 2022 Chang S, Greenwald B, Roy K. Published by Cell and Gene Therapy Insights under Creative Commons License Deed CC BY NC ND 4.0.

Article source: This article is a transcript of a webinar, which can be found here.

Webinar recorded: Feb 03 2022; Revised manuscript received: Mar 15 2022; Publication date: Apr 22 2022.