Please can you introduce yourself?
SM: Hi I’m Sigma Mostafa vice president for Upstream and Downstream process development at KBI biopharma. We are a CDMO (Contract Development and Manufacturing Organisation) and we have two locations in North Carolina, as well as in Colorado.
What is process robustness?
SM: The general definition of robustness is the ability of a process to tolerate the expected variabilities, and this comes from raw materials or process conditions or human factors or equipment. Essentially, robustness is something we need to demonstrate prior to a molecule going into market. A lot of the process characterisation work gets done later in the life cycle of a molecule to ensure that data is generated. So what we are talking about here is bringing in the sort of studies of robustness as well as understanding around the cell line process very early on so that there aren’t any surprises in the late phases of the molecule.
What are some strategies for ensuring process robustness?
SM: In general with a QBD approach, there is that idea of building quality into the process, so from the very early stage of the brief, to clinical trials, to early-stage clinical trials we are definitely trying to make sure that the design of the studies that are being done are through a proper statistical design.
We collect data early through late phase to build in robustness throughout the lifecycle. Some of the things that help in that regard are some of the high-throughput technologies which allows us to look at a lot of different parameters, both in terms of the process as well as the analytics part. This is because there are high throughput technologies now in the analytical side which allows us to basically get a lot of data, and not compromise on the data amount, as we are trying to move molecules faster through the early phase of development in phase one or so, because oftentimes the other approach that has been done is platform development.
What that means is you have built a platform for upstream and downstream development, and you essentially put all your molecules through that particular platform. But as a result, you get very limited information, you do not know how your cell line or your particular product reacts to different process parameter conditions.
So by using high-throughput technologies such as ambr 250 bioreactors, they can be used for some process characterisation work as well as early stage development. This allows for essentially gaining a lot more information, so you don’t lose out on that information.
What are the shortcomings of current cell lines to meet future/current needs?
SM: So cell lines in general, the technologies around (I think vector design) as well as the host designs have come a long way. That has allowed us to achieve these very high titres that we now see and often the discussion is ok now that upstream has achieved this 10g per litre (at least) in fed batch and maybe 30g per litre when you’re talking about high cell density processes, the challenge is now on the downstream.
However having said that, the focuses really still on productivity and not so much in terms of getting some other independent parameters that you can look at early on for a cell line to know whether it’s going to be a robust cell line, and will be able to withstand the process variabilities that you can expect as you go to larger scales or you go to different size etc.
So I think more understanding of CHO cell lines and being able to characterise them more. This certainly helps, and there certainly will be proteomics work to be done. Still I don’t think the understanding of product quality in terms of how specific cell lines or clones affect product quality or this concept of cell line robustness have not been studied further. So I think there’s still some improvement there that can be done.
What are the drivers for your work?
SM: So we are of course a CDMO so we work on fifty or so molecules per year, and a lot of those molecules are actually where we are developing the cell lines in house. The context here was that these were cell lines that we had received from certain clients and in choosing those final cell lines the driver in most cases has been productivity and not a lot of thought around how this alignment perform under different process conditions. So our driver was essentially to understand more how we can early on pick a cell line that will make the later phases of the work including the characterisation studies easier and we are not going to have a such a situation where the normal operating range and the proven acceptable range are very close to each other because basically if you go slightly away from certain pH range or slightly away from a certain temperature range due to your cell line characteristics it will fail.
So obviously, for mammalian cell lines still there is a sense of it’s more of an art. We don’t fully understand all the attributes of a cell line and what controls that. But again, building in some of that understanding is what we want to be able to do and we are looking forward to collaborating with some groups both academic and industrial to achieve that goal.