The variable nature of advanced therapies means that a high level of consistency of raw materials is a critical success factor. Comparing raw materials providers, measuring lot-to-lot consistency and accounting for variations can be arduous and take focus away from other scale-up and commercialisation efforts.

So, what are the best practice approaches when it comes to measuring cytokine activity and function and setting risk-based parameters for variability?

In this interview, Kate Fynes, CMC Translation Consultant at eXmoor Pharma Concepts shares her views and experiences.

CB: Are international units a help or a hindrance?

KF: Being of biological origin and therefore prone to variability, international units (IU) provide a guide to the biological activity of each cytokine lot. This should enable material from a new lot or supplier to be used interchangeably between runs whilst maintaining process control.

In practise, however, when considering IU, caution must be exercised. Establishment of IU is based on biological assays which are inherently variable. Units should be considered a benchmark of bioactivity rather than an exact value. Depending on the downstream application, on receipt of a new lot it may be prudent for the user to conduct an additional biological assay in-house, adjusted for their specific application. This could involve running a simple cell culture expansion curve against internal reference standards or testing the new lot in a well-established small-scale process to ensure consistent product quality within validated ranges.

CB: Is meaningful activity and function testing possible without standardised units across all suppliers? 

KF: The specific activity of a cytokine measured in a different cell line and cultivated under different conditions in a different lab by a different operator can differ significantly. Without standardised units across suppliers, you lose the ‘activity benchmark’ of a new lot when switching. Therefore, more broad characterisation data would need to be generated to identify the optimal unit ranges required to maintain critical quality attributes (CQAs) and thus, product quality. This is not to say meaningful activity and functional testing cannot be obtained without standardised units, but the time and money required to perform a greater number of characterisation studies could be considerable, especially if not considered in advance as part of your development strategy. 

During my time in process development, I have experienced first-hand the challenges of unplanned supplier changes and the complex comparability packages that were required as a result. These changes resulted from both necessary platform optimisation and unexpected supply chain issues, both of which are not uncommon during development. My advice would be to establish strategies for bridging lot-to-lot and supplier-to-supplier variability early in development and, if relevant, consider establishing in-house assays to determine the specific activity of incoming material to allow standardisation at the developer end.  

CB: Can you tell us more about your approach to measuring cytokine bioactivity and managing any variability between lots? 

KF: During optimisation of different cell processes, I have evaluated cytokines from multiple suppliers. Some are able to manufacture material of a consistent lot-to-lot activity, whilst others do not have this capability. In the case of the latter, the specific activity of the lot must be referenced each time to calculate the required resuspension volume and achieve a desired concentration. 

This is time-consuming and has the potential for operator error during an already long and involved processing session. This way of working is not GMP-appropriate where the need for standardised process steps is even greater. As mentioned, I have been involved in a change in cytokine supplier to enable standardised reconstitution practices that ensure lot-to-lot consistency as part of the GMP transition. This cost us financially and extended our timelines to clinic. 

In terms of measuring incoming cytokine activity, IU was typically used as a benchmark in the process development activities I have been involved in. This was sufficient, as the characterisation data around target concentration indicated a wide process acceptable range (PAR) that could still achieve the desired product quality. Therefore, in these scenarios, a further in-house assay wasn’t required. I would highlight here how critical it is to build time into your development plan to characterise and understand the process as this will enable definition of a robust control strategy. This will then allow you to anticipate and pragmatically respond to a deviation in an ancillary material and how it will affect product quality.  

CB: How do you work with raw material suppliers to ensure lot-to-lot consistency and avoid comparability studies? 

KF: In response to customer need, most suppliers are now able to release product in a consistent manner. However, as part of manufacturing a medicinal product, it is good practice to audit suppliers of critical raw materials and ensure the product works for you. Be that from a quality, format, supply, or stability point of view, for example. As part of an audit of a cytokine supplier, we would assess the credibility and validity of their testing strategy against IU specifications and the subsequent release strategy. It is also common practice to establish a supply agreement with the supplier once you lock down on a process reagent to ensure your supply is protected and any modifications to the reagent are not made without clear communication. If it is not possible to ensure lot-to-lot consistency, depending on your dosing schedule requirements it could be an option to reserve an entire lot that may support your clinical study program and into commercial, with comparability only required once the current lot runs out. This will minimise the comparability efforts, but the downside could be storage of a large lot of reagents. 

Looking back, I have certainly taken learnings from when proactive material management has not been executed quite so well! On one occasion, translation to GMP was considered too late in process development and having obtained considerable data with a single cytokine supplier (error), it was confirmed that the supplier could not release consistent lots. This was an important criterion to enable a standardised resuspension step in the GMP manufacturing process. As a result, we established two parallel approaches to mitigate risk that cost us time and effort:

  1. Studies were executed to evaluate other suppliers who could release material with an acceptable level of lot-to-lot consistency. Once comparability was demonstrated with a number of suppliers, a process change was made to include multiple suppliers on the approved supplier list to mitigate future risk. The comparability package was designed such that it allowed us to bridge the development data already acquired removing the need to repeat the studies with the new cytokine source. 
  2. The process was further characterised using the original cytokine source to better define the control strategy around that reagent. A greater understanding of cytokine concentration as a critical process parameter (CPP) and how this affected CQAs enabled a wider range of cytokine activity to be accepted. This expanded PAR encompassed the supplier’s validated release specification range. As a result, the process could be streamlined and simplified, each time resuspending one whole vial in a set volume rather than calculating a required volume and subsequently inputting a fraction of the vial. This confirmed that any variability in the specific lot activity would still be within the established PAR and would not detrimentally affect CQAs and therefore product quality.


CB: Will comparability studies ever be completely avoidable? 

KF: In short – no! Even with well-defined development strategies and proactive planning, changes are inevitable on the path to the commercial supply of a medicine. A manufacturer will always require demonstration of comparability in terms of impact to product quality. Changes can take the form of anything from a simple extra wash step; a new reagent supplier; a shortened process, a shift from a manual process to a fully automated platform; through to the ultimate comparability with a change in the manufacturing facility. Comparability studies should be viewed as an inevitable and essential tool to enable process change. Proactive establishment of a clear development strategy that will help anticipate inevitable changes and their timelines, as well as defining a phase-appropriate control strategy, should minimise and will aid in the execution of unavoidable comparability studies. It’s an aspect of my role at eXmoor that I enjoy most – saving clients time and money as they progress through the development of their products. 

About Kate Fynes:

Kate has 15 years’ experience in cell & gene therapy research, process development and GMP translation within academic, biotech and big pharma settings. She has specific expertise in closure, automation, scale-up and tech transfer of cell therapy processes, as well as in applying a risk-based approach to the development of robust, phase-appropriate control strategies.