Comparability for advanced therapy medicinal products with Christopher Bravery

For a recent KNect365 webinar, Christopher Bravery from Advanced Biologicals discussed the challenges of comparability for the ‘more complex’ biological molecules. He went into a detailed examination of purification, characterisation and activation, and their position within the space of biological comparability. Here we summarise the in-depth discussion that offers guidance and advice for those looking to meet the requirements of comparability for their molecules. 


First, Bravery outlines the differences between small molecules and biologicals which makes the characterisation of them a greater challenge for pharmaceutical companies. For small molecules, it is possible to know the exact molecular structure. When a cell substrate is introduced, the ‘manufacturing process is less clear’ as the ‘windows of characterisation are fogged with variability.’ This makes comparability much more challenging for cell based products as you are unable to define the molecular structure as easily. The proteins and the cell itself all vary in their phenotype, and are often more heterogenous, which adds to the uncertainty in the measurements taken.

Bravery defines comparability for biologics as the ‘need for a change in the process’ when the development is incomplete, either at post-approval or late stage. At this point, you should have characterised the raw materials to ‘understand critical quality attributes to establish process parameters’. He uses a hypothetical case study for a CAR-T product to demonstrate where comparability can need to be used within a development system, highlighting the area of ‘activation.’ (see diagram top)

Bravery begins by outlining how to ‘characterise and develop the tools that would allow you to undertake a process change at a later point.’ The cells entering the activation step have already been purified, and are a critical intermediate in the step. Bravery states ‘To be able to set the specification for this step, you need to do some characterisation, ideally early in development.’ This means looking at the phenotype of the cell, including, the purity, impurity and markers that have been activated in the cell. There needs to be an awareness of how many markers are activated at the beginning and the end of the activation step; this will help with identifying characterisation and is conducive to allowing for comparability later in the process.

The ‘Characterisation of Unit Operations’ is vital to allow for comparability. Bravery notes if this has not been previously considered, altering the process becomes a challenge and more time consuming later. ‘Characterisation should use both orthogonal and alternative methods’, with Bravery describing ‘flow cytometry’ as a natural choice for the characterisation for T Cells, which he uses as an example. He divides his unit operations into three sections: the analytes which are the markers, the measurand which is what you are trying to measure, and the purpose, which is what the measurand is characterised as. Bravery uses the example of CD25 which is an IL2 receptor due to the activation step. Cells that are activated with IL2 will be CD25 positive, with cells that are not activated as CD25 negative. Intercellular straining is another option for IL2 indication if you are using IL2 for characterisation. Bravery warns however, that this might not be completely successful as intercellular straining is far more complicated.

‘One of the principles in characterisation for any product is to use orthogonal methods, which is looking at the same attribute by a different measurement principle.’ If IL2 is important, you can also perform an ELISA, which is a measurement of the secretion of IL2 into the medium. However, the protein methods might not be the best for the characterisation of the cells, so a PCR method can also be used.

The intercellular straining, ELISA and PCR are all orthogonal methods of measuring activation, with the flow cytometry an alternative method where you measure the receptor as opposed to the ligand. In choosing any of these methods, if it is useful it can become part of an extended tool kit for characterisation ‘for later comparability exercises.’

Bravery goes onto to give learning points derived from his activation characterisation, stating:

‘Your characterisation aims to identify the critical and other quality attributes. You need to understand those in order to explore the process parameters. This gives you an understanding of your overall product and process and allows you to consider doing comparability exercises…It is important for the characterisation to use more than one attribute and possibly look at alternatives methods for the attribute, or different aspects of the attribute.’

Bravery returns to the original flow diagram to demonstrate how the critical parameters can now be understood once characterisation has been completed. The process parameters now need to be examined, for example, temperature, CO2 concentration and duration, all of which will indicate an optimised process. The measurement of the process parameters can be done ‘parameter by parameter, where you can measure the effect of the parameter on the quality attribute. Bravery states ‘you need to understand the criticality of each process parameter to establish a normal operating range (NOR).’ All data gathered is critical to justify the operational ranges for these process parameters.

After characterisation has been done, it is at this point when a comparability exercise may need to be undertaken if part of the process needs to be altered. Bravery uses the example of needing to source an alternative bead supplier. The comparability exercise should aim to actively look for differences, not just to show the products are directly comparable, which is why orthogonal methods and stability methods, which are sensitive to change, are the most effective characterisations.

Although these changes may seem small and inconsequential, Bravery stresses the changes can affect the ‘quality, safety, and efficacy of the product’, and that each attribute should be considered separately. It can also often include a ‘revaluation of stability to demonstrate the change does not alter the product.’

To prove that this change is not altering the product, first, all controllable sources of variability need to be eliminated. This includes: the same starting material, the same raw material and the same test samples. By doing this it demonstrates none of these variables will alter the product. It removes any ‘noise’ within the various results. Bravery’s main message is that ‘specifications are a sub-set of what is required to fully characterise, the starting/raw materials, intermediates, drug substances (where declared), drug product, and process.’ Any change to these aspects of the process will need to be recharacterized to demonstrate comparability in a product.

Bravery also outlines some common comparability mistakes such as not meeting the existing specifications, as a small change in quality can have unexpected clinical outcomes. ‘Comparability is determined by the totality of the data.’ He states that avoiding the word ‘comparable’ in the data is important due the difference of its meaning in the regulatory sense, suggesting the use of ‘highly similar’, or ‘no significant difference.’

Bravery concludes by highlighting the key conclusions, which includes the need to plan well, to do extended characterisations, include stability testing, and maintain an awareness for clinical and non-clinical data. A product is only comparable when you prove there is ‘no significant impact on safety or efficacy derived from the change. If these requirements are acknowledged and understood, it will allow for a more successful comparability exercise, which will reduce process time.

Listen to the webinar and explore seven other topic sessions on the Cell and Gene Therapy Digital Week hub.

Leave a Reply