The Real Value Of Continuous Bioprocessing Is In Connected Unit Operations

By Tyler Menichiello, Chief Editor, Bioprocess Online

AstraZeneca (AZ) is leading the pack when it comes to innovation in continuous biomanufacturing. Perhaps I’m just biased, but at the very least, they seem to be the most vocal proponents of continuous biomanufacturing, and that doesn’t appear to be changing anytime soon.

 At this year’s BioProcess International conference, AZ’s executive director of bioprocess technology and engineering, Jon Coffman, Ph.D., delivered a terrific plenary keynote titled “Integrated And Continuous Bioprocessing: What’s Holding Us Back, And Where Do We Go From Here?”

His talk offered a compelling case for continuous systems, which they refer to at AZ as “Integrated Dynamic Bioprocessing.”

“The word ‘continuous’ has fallen out of favor,” Coffman told the audience. Though he didn’t quite expound upon why, integrated dynamic bioprocessing (IDB) is a more appropriate catch-all term for the type of production Coffman was describing.

“We like the idea that it’s integrated,” said Coffman, “and that may be the most important thing that we’re doing.”

After all, the integration is where the value is. “Connecting and automating unit operations together has tremendous benefit,” said Coffman. “Eighty percent of the value comes from making it connected.”

The “dynamic” element refers to the fact that AZ’s integrated approach is “not steady state,” but rather, an increasingly productive and fully connected process.

“We’re developing very short continuous processes [14 days], and we’re doing a very high cell density,” Coffman explained. “The productivity is initially three times better than fed-batch, but we can actually increase that during a later stage.” This results in higher cell density and higher productivity in a shorter amount of time compared to traditional fed-batch or even steady-state perfusion systems.

During his keynote, Coffman described the cost-saving implications of IDB, as well as the unique engineering challenges it presents, and he emphatically supported a broader shift in the adoption of continuous and integrated systems.

“We should never build another plant that doesn't have connected processing downstream,” he told the audience.

Tailoring The Value Proposition

There’s been a business case for continuous manufacturing in pharma for some time, but biologics have lagged behind due to their unique technical and economic complexity. However, companies like AZ are showing that its time may be here for biopharma.

IDB is useful to biopharma organizations everywhere, argued Coffman, from use in R&D to commercial operations. While use cases may change across the drug development continuum, the ultimate benefit remains: Integrated and continuous bioprocessing offers significant cost savings.

Of course, the case for IDB very much depends on who you’re pitching it to.

R&D teams may not care about the cost of getting a product to clinical trials, which they aim to do as quickly as possible, but they’re certainly interested in saving their budgets, said Coffman.

“On the other hand, commercial operations is really interested in [IDB], because it can reduce the cost of goods and really reduce the CapEx for building a new plant,” he told the audience.

Understandably, the cost savings of IDB are more pronounced at the commercial scale. Cost-sensitive and high-volume products represent a significant opportunity for companies like AZ to save “potentially billions” through IDB, according to Coffman.

Maintaining Continuous Process Control

The nature of continuous bioprocessing demands a higher degree of analytics and process control to keep things in spec. Describing the challenges of monitoring cells in a bioreactor, Coffman used the analogy of people stuffed inside a phone booth.

“The number of cells that we have is about the same as putting 10 people in that [~ 2,000 L] phone booth,” he said. “Basically, we’re trying to feed 10 people for 14 to 20 days, so we need to figure out how to feed them; we need to figure out how to give them oxygen; we need to figure out how to remove carbon dioxide; and of course, there’s a waste problem that we have to figure out as well.”

This, on top of having to remove the product itself, presents a substantial need for both real-time analytics and automation.

AZ uses both digital twins and process analytical technologies (PATs) to increase process robustness and control, Coffman told the audience. (Another session by AZ’s director of MSAT process analytics and laboratories explored the company’s focus on predictive control and semi-autonomous processing.)

Raman spectroscopy is used to measure glucose, said Coffman, and that data then feeds into a model to predict titer. “We have a mechanistic model that relates to the cell density and cell viability, and then estimating the particle count to filter out,” he explained. “Then, we can use that to show how much sitting there is, how much the titer goes down when it goes to the TFF system, and then estimate how much we put on the protein column or affinity column.”

The Engineering Challenges Of IDB

“It turns out, when you do integrated continuous [manufacturing], you are incidentally making a drug, but you’re really making buffers and media,” Coffman half-joked. “Your plant for making buffers and media can be bigger than your floor plan unless you plan it correctly,” he told the audience.

Buffer concentrates and inline buffer formulation are critical aspects of IDB, said Coffman, as well media concentrates. “We would be a media prep plant if we didn’t have media concentrates,” he explained.

Perfusion, arguably the most challenging step in any continuous process, relies on effective tangential flow filtration (TFF). For a 5,000 L bioreactor, the combination of necessary pumps and filters would have been a “nightmare,” said Coffman.

“We knew that was not possible, so from the very beginning, we’ve been developing methods to make those systems run longer and clean up the cell culture so that it doesn’t foul filters as much.”

So, what was the single greatest advance AZ made to that effect? Simply cleaning the filter.

“We flush it with media; we clean it with hydroxide; we clean it with media again; and we put it back into duty,” Coffman explained.

“This is maybe one of the things I’m most proud of in terms of what we’ve done,” he told the audience. Regularly cleaning the filter allowed AZ to scale up to 10,000 L “without a problem.”

Coffman described initial pushback and skepticism by AZ’s upstream scientists as “severely negative,” and understandably so. Filter cleaning is not uncommon downstream, but it’s almost unheard of upstream, and they were legitimately concerned about the cells being exposed to contaminants. These were “easy engineering things to work on,” Coffman said, but it still took months to convince AZ’s upstream scientists of the benefit.

“The food industry does this all the time,” Coffman explained. “That’s why I’m proud of this, because it took something from a different industry, brought it in, and made it routine for us. Now, we have automated systems that will do this easier than changing the filter.”

Despite the unique considerations and engineering challenges IDB or continuous manufacturing presents, my impression is that it’s worth it.

As Coffman emphasized throughout his keynote, “Making [a process] continuous is not as valuable as making it connected.” He urged those in attendance to look at their current processes and see what they can do to link them.

“This is like 14 years of work, tens of millions of dollars in terms of development and practice,” Coffman said in closing. “You don’t have to do that. You can learn from other people and other companies and develop your own process.”