Advanced Therapy Medicinal Products (ATMPs), or Cell and Gene Therapies (CGT), have the incredible potential to cure devastating illnesses, such as cancer, on a more personalized level. But, due to the production variability of personalized medicines, these processes are some of the hardest to control.
When creating these products, especially in the research phase, a huge amount of useful data is produced every day. The abundance of data across variable processes can challenge data management principles and risk data integrity.
While the processes may not be easy to control, with the proper data integrity plan, the data can be controlled and applied to improve these innovative therapies.
Generating Variable Data from Variable Processes
When designing personalized therapies, different methods can be used to determine the best cure for a specific patient.
For example, consider the process of Autologous Cell Therapy, where blood or tissue is taken from the patient and is used as a starting material for isolation of the patient’s T-cells that best recognize cancer cells. These specific T-cells are then used as a starting material for further upscaling and processing. Finally, the personalized T-cells are given back to the patient. In this process, we are dealing with the intrinsic variability of the starting material, the cell growth rate, and the cells behaviour under in vitro culturing conditions.
A different method is to modify the T-cells by introducing a transgene encoding a chimeric antigen receptor (CAR). This modification is usually performed with a lentiviral vector. Now, the variability of the receptor is dependent on the ex-vivo modification process. Because we are dealing with DNA technology and living material used in production processes, the outcome is not always 100% predictable. And this is only the beginning. It gets even more complicated when the modified T-cells are scaled-up. We know that every culture will deal with different genetic background from both T-cell starting material, receptor variability, and the final product. This simply means a lot of (variable) data is generated throughout the process.
GxP principles require controlled and reproducible production processes to end up with a reproducible product. Setting standardized controls to ensure a reproducible process is challenging for ATMPs, because starting material changes for every patient, the process parameters are different for every production process, and the final product is different for every patient.
Regulatory authorities expect us to be in control of production processes and show reproducible results. So, how do we show that we are in control of our processes and deliver the product that we indicated to deliver? The only way to show we are in control is to present the applicable generated data.
Understanding Data Management Principles
Handling a great deal of changing data is an easy way for something to get lost in translation. Therefore, we are sharing some ideas to help get things under control and ensure data integrity.
Document, Document, Document…
The first step is often skipped but is the most important. Start documenting early, even in the research phase. While developing the ATMP, you may be excited to see the first promising results, but it is critical to document all experiments, materials, and methods carefully. Don’t throw out data from experiments that were not successful, they might be useful later. Make sure you are always able to look back at the data being generated at these steps of the process, make sure it is findable. Although the data might not all be used for the final product, the data might still be valuable for further investigation or future root cause analyses.
Sort It Out
From a well-documented and understood early stage development process, it is possible to develop a reliable and robust manufacturing process. A lot of effort is put into the development of upscaling physical production process, but at the same time it is important not to underestimate compliant data management.
Once you have insight into the process and fully understand what kind of data is produced, you need to realistically identify which parts of the processes you can control and decide what data you need to demonstrate is in control. This is both the most specific and difficult part of ATMP development.
A few simple questions will help you define when data is important to your process and needs to demonstrate control.
- Where in the process flow is data generated?
- What quality decisions do you (or someone else) make with this data, if any? Can the data be associated with any Critical Quality Attribute? What do you need this data for?
- Is the data generated from this process raw data or is it processed by any means?
- Will you be able to reconstruct the process in five years?
- Which are the scientific rationales behind the decisions that have been taken?
Having applied the questions above, you will be able to gain understanding of which critical data to manage and, with a risk assessment, it will come to light where to set any data controls.
Setting Data Controls
You have systems in place for the production processes and know your critical data, now you will need to set the right checks to demonstrate product data and in-process controls are trustworthy. Setting the process controls for ATMPs might be challenging, but dealing with the data from the process should be easy if you know the basics.
There are three areas of Data Integrity control setting, by:
- Behavioral controls
- Technical controls
- Procedural controls
Each of these controls can ensure the data is kept complete following the ALCOA+ (Attributable, Legible, Contemporaneous, Original, Accurate, + Complete, Consistent, Enduring, Available) principles. Most often, different types of controls need to be combined; for example, if procedures do not describe where to store data, the technical control on its own would fail from having an access-controlled data folder.
In the world of ATMPs, an agile manufacturing process with an automated process control that adapts to the variabilities would be a great technical solution if consistent quality and effectiveness of its products can be shown. But automated systems are only as good as how they are used. Therefore, we also need behavioral and procedural controls. Only the right combination of control settings will give the data integrity control needed for ATMP production.
Benefitting from Reliable Data
In the earlier case of Autologous Cell Therapy, production processes might not be easy to control, but several data integrity controls can be applied to make sure your data can support your business.
It is true that processes under GxP are expensive, requiring time and resources due to more ‘administrative activities’, demonstrating compliance, and the requirement to continuously improve. However, good data management, including mature data integrity, can save time and money in the end.
When a problem occurs, good data management backs up your solution to solve the problem. Traceable, findable, and integer data represents business value (‘what can you do with all that information?’). Business decisions can only be made with trustworthy data because teams need to backtrack to the source of the issue. Often, this can be the difference between winning and losing. Winners can recover while others lose the built value and need to start over.
If you need support setting up good data management practices for innovative therapies, partner with ProPharma Group. As a global provider of end-to-end cell and gene therapy solutions, as well as Data Integrity services, our Center of Excellence accelerates the commercialization of advanced therapeutic products while ensuring compliance with all Data Integrity requirements. Contact us today to learn how we can help you overcome scientific, technical, and regulatory challenges within the cell and gene therapy space.
