Will downstream productivity catch up with the upstream in 2024?

Industry will see mid-teens grams/L of product within five years but downstream processes are the big ‘virgin territory’ for further productivity gains, says Avantor’s Ger Brophy.

In the aftermath of CPHI Barcelona – the world’s largest pharmaceutical event, where biologic equipment, innovators, and CDMOs gathered in October – we spoke with Ger Brophy, outgoing EVP of Biopharma Production at Avantor and incoming chair of Avantor’s Scientific Advisory Board, on his predictions for improvements in bioproduction efficiency.

Brophy, one of the experts who contributed to the CPHI Annual Report, predicted that over the next 24-months the industry will implement incremental process progress, driven primarily by new biologics license applications (BLAs), with older approvals unlikely to see process adaption. However, looking ahead five years, the impressive product yields (circa 14-16 grams/L) we presently only see at conferences from products delivered at smaller scales will be in much wider industrial use.

Additionally, downstream process efficiencies are the next pressure point for industry to target, having already made good strides in many upstream areas. Regarding advanced therapies, process improvements are likely to be more muted as innovators rightly settle for bringing these vital therapies to market first. Longer term, he predicts step change improvements similar to those already seen with monoclonal antibodies (mAbs) during the last two decades.

BPI: What process improvements might we see in the next 24-months to help bring down the cost of manufacturing, particularly in the downstream?

GB: The first thing to make clear is that 24-months in our highly regulated industry is perhaps not as long as one might think, as ultimately the process is the product, and we need to understand all the downstream consequences. However, improvements are happening continually. It’s just that significant improvements may take a little longer, maybe five or seven years. Perhaps the best place to start is to look at what we have already seen and, overall, most of the biggest and most successful improvements in productivity to date have been in the upstream processes (like fermentation etc.).  These have already delivered step change improvements in yields. Fast forward to today, and we see many companies at conferences even presenting 15 to 16 grams – up from a couple of hundred milligrams back in the 1990s.

While we have moved steadily forward in upstream processes, I think we can now expect the same level of productivity improvements to start emerging from downstream processes, where to date we have not made as significant of productivity gains.

For example, to take the protein A capture step: We might see small initial and incremental improvements using fibre capture instead of chromatography, and we now see researchers also looking at surface area and flow through characteristics. Then if we look at process intensification improvements, we can identify a few that are going to help lower the total cost of ownership.

For a specific prediction, my crystal ball says that for processes we see industrially – not just in the lab – we may see regular new BLA processes in 5-years’ time that achieve the mid-teens per gram. But it’s also important to say these will be process dependent and high single digits, even five years from now and would be a dramatic improvement for certain products and already a 2-4x improvement.

BPI: Continuous bioprocess still appears in its earliest stages of implementation – what developments do you envisage here over the next five years?

GB: There has been a lot of talk around this, and we have to realize that in any industry we very rarely see end-to-end continuous processes, and in pharmaceuticals to be able to define a batch is often critical. It’s easier from a regulatory point of view to have a discrete batch [especially when there are problems].

We also have to accept that with a lot of biologics, we don’t yet have full visibility on the molecular processes, so in the near future, we are going to look only at individual process. For instance, we’ll need to look at continuous chromatography, protein A chromatography and fermentation, then perhaps we will explore if we can link these subroutines together.

BPI: Over the next few years’ it is predicted that biological stakeholders will increasingly partner to lower the costs of production – as a provider of manufacturing services, how can companies like Avantor help share manufacturing improvements?

GB: I think it’s partly driven by the fact that perhaps pharma companies don’t necessarily need to be great at biopharma process improvements and manufacturing to be successful at bringing a drug to market. So what you will see is companies like Avantor and the CMOs and CDMOs driving strategic learnings between them about what works. For example, when a smaller biotech comes into early development, the CMOs might be able to get them up and running within just a few weeks or months using an almost plug-and-play method adapted from processes that have worked before.

For advanced therapies, the story is a little different. While the clinical benefit is undeniable, in terms of the process and the cost of these products then yes, we certainly have some way still to go. The interesting thing moving forwards is that we see much more sharing now between manufacturing technologies companies and CDMOs, and maybe even academia to scale new processes. This is potentially a great sign for accelerated innovation.”

BPI: We have heard a lot of talk this year about AI helping optimize bioprocesses in the years ahead – what role will it play?

GB: The digitalization of processes and process characterization will obviously continue apace, but AI won’t be the only game in town. We see lots of efforts in moving forwards with areas like ‘digital twins’. The next steps are to build the large data lakes that AI platforms need to undertake its best work. I think it will be part of the picture, but other digitization efforts and characterization efforts in pharma 4.0 will be more significant.

BPI:  Looking specifically at cell and gene therapies, what innovations and optimization do you anticipate in the next two to five years potentially coming from academia?

GB: We may see mid-teen approvals in advanced therapies this year and next, and there is no more exciting space out there as it’s curing – not simply treating – disease. That being said, it still takes a significant amount of material to make a dose, and autologous therapies by nature are slow and inefficient. I think we will see incremental improvements, and what I anticipate happening is manufacturing technology won’t be able to keep pace with the clinical speed of development.

Monoclonal antibodies maybe a good yard stick here – if not exactly like-for-like – so it has taken roughly two decades to see 2-3x improvements. I think we are looking at the next decade to see bigger cost improvements – but it will happen, because it has to happen, so we can democratize therapies.

BPI: How could these processes improve future commercial therapies (and by when) – are any of them close to being used in an approved therapy?

GB: I do anticipate there will be significant improvements in finishing chromatography for things like viral vectors and a number of companies are working on technological innovations. The important point to stress is that any improvement in process intensity is going to pay dividends, as these therapies remain very expensive. We will see improvements in process-finished chromatography and filtration for a number of these new modalities, as some of the existing purification technologies like caesium chloride are still difficult to scale, but these will all be for new BLAs rather than improving the efficiency of existing products.

BPI: Are regulators now looking more at helping improve speed or development?

GB: Both. The triggering event was undoubtedly the Covid pandemic, a period when we suddenly had – because we had to – a re-evaluation of what can be done safely and by when. The dust is now starting to settle, but from the regulators’ point of view, there is an understanding that we won’t go back to business as usual. As an industry, we are now looking at what are the learnings from this and what we can keep to improve all development programs moving forwards. Hopefully, this can be a lasting positive from this post-pandemic area – faster medicine approvals really would be a wonderful legacy to have.

Bioprocessing efficiency was a central cornerstone of the onsite debate at CPHI Barcelona, including via a dedicated Next-Gen Biomanufacturing track that can be watched back online.