Continuous processing is pretty much a “given” in many industries — even the larger pharmaceutical industry that makes synthetic small-molecule drugs. But the concept has only just begun to make inroads with biomanufacturers, who have until recently worked mainly in batch or fed-batch mode. Single-use technologies largely have enabled them to consider the possibility of process intensification and going continuous. In support of this month’s featured report, I asked contributor Margit Holzer, PhD (scientific director at Ulysse Consult S.a.r.L in Luxembourg) a few general questions on the topic.
One question often asked of perfusion and other continuous culture approaches is “How do you define a batch?” Some experts say it isn’t even necessary to do so. Others say you can describe batches in terms of time. What do you think?
HOLZER: It is important to recall that traceability of the whole batch history of a drug substance or product is absolutely necessary for CGMP production. So there is no choice. A batch also needs to be defined for continuous upstream operations. This becomes especially crucial during investigations or product recalls. In the case of continuous production, a batch may correspond to a defined fraction of production.
The batch size of a continuous upstream process can be defined, for example, by a fixed quantity (e.g., volume, mass, or activity units of product) of harvested product; the amount produced within a fixed time interval (e.g., hours of production, residence time) between harvests; or the number of cell generations or doubling times to be produced, collected, and further treated in downstream processing as one batch. In addition, a minimum titer and/or viability and/or other quality requirements can be specified as acceptance criteria for pooling with the harvested product to assure batch homogeneity. For all those cases, downstream processing capacity must be in line with the harvested quantity of material.
What’s the most challenging part(s) of downstream processing to do continuously?
HOLZER: Many technologies supporting continuous downstream processing have been developed (or are being developed) during the past 10 years. One of the most important is continuous chromatography, which is for a great majority of applications probably the most challenging concept.
Many case studies demonstrate that continuous monoclonal antibody capturing using protein A affinity chromatography can be run reliably over hundreds of cycles. However, applying continuous chromatography on mixed-mode, ion-exchange, hydrophobic-interaction, and linear-gradient chromatography is more challenging and needs more process characterization and experience. For instance, definition of product residence time (in case of recycling of side fractions) becomes important. Another task that needs to be addressed in this context is development of models for viral-removal studies. Generally, process control and management of out-of-trend or -control situations need to be thought through differently than for typical batch operations.
Just to be clear, economically it does not make sense to transform all unit operations that are run in batch mode into continuous mode. For example, 0.2-µm filtration steps can be run easily in an alternating–batch mode, switching from one to the other filter.
How are disposables enabling continuous processing overall? Much of what we’ve published on the topic seems to indicate that they are vital to making it work.
HOLZER: I would rather make the inverse statement: that continuous processes allow us to shrink process lines and therefore make possible the implementation of single-use technologies at manufacturing scale (both technically and economically).
Consider a protein A capture step. Typically, production scale uses columns with an internal diameter of 60–100 cm and protein A volumes of 50–150 L. No single-use prepacked CGMP columns are available currently at that scale. Furthermore, even if such disposable columns were available, and we used up to 30 cycles during a campaign, the cost contribution of just the protein A media would be extremely high (typically >650,000 €). By transforming that batch capture process into a continuous one, we could work on two factors to shrink column size: We can at least double column capture capacity, often by running chromatography in countercurrent mode, and requiring only half of the packed-bed volume. We can run the operation faster and 24 hours/day to achieve much higher cycle numbers within the processing time (again shrinking the necessary column size). Last, but not least, continuous chromatography processes need several columns. In this case study, the operation could run with four 2-L to 8-L columns (ID 15–30 cm). Such dimensions allow us to implement single-use technology.
Requirements for batch and continuous processing applications are different for single-use technology because continuous operation typically uses the technology over a longer period that might also include cleaning cycles for certain downstream unit operations. So material requirements for such disposable technologies run in continuous mode are more demanding to keep integrity and functionality over their period of use.
Finally, I really think that adapted disposable technologies can help a lot to make it easier to implement continuous contained processing lines.
Some people talk about “blurring the lines” between upstream and downstream — presumably to create a seamless, integrated biomanufacturing process that can occur in one single ballroom suite using closed systems that are connected from bioreactor through purification of final drug substance. Can you speak to the contamination risks associated with this idea and whether you believe it will be possible?
HOLZER: Bringing up- and downstream closer together — in terms of process integration and operations — is absolutely necessary for integrated continuous biomanufacturing. However, this does not necessarily mean working in a single ballroom. As for batch operations, risk assessments need to evaluate and treat the risk of cross-contamination, viral contamination, microbial contamination, mix-up, product flow, and so on. Even with closed systems, each single operation (e.g., sampling, room cleaning, and decontamination procedures) must be evaluated for its impact (e.g., of leakage of connections or breakdown of tubing) on all operations going on in the ballroom.
The risk level depends on several factors: whether animal-derived components are used in a process, the nature of the production cell line itself, whether the production area is dedicated to a single product or part of a multiproduct operation, the overall process duration, and ballroom containment. All together, many risks need to be assessed for the individual product and process, and acceptable response plans should be developed. However, before evaluating continuous integrated processing of up- and downstream in a single ballroom, we need to make sure that a process runs reproducibly under controlled conditions.
I can’t imagine drug-product formulation and fill–finish getting physically connected to that integrated drug-substance process line described above. Isn’t that asking too much?
It really depends on the motivation to run a whole process from drug substance to drug product continuously. Could there be some specific supply chain needs, for example, or is a product very unstable? Product hold steps at the drug-substance stage provide flexibility to adjust production schedules according to different drug-product needs. Furthermore, fill–finish operations often are performed by sites and companies with specific know-how and skills. In terms of technology, it’s not so challenging to transform batch formulation and fill–finish operations into continuous ones; most filling and packaging lines already run continuously. So it only would need a focus on formulation.
Generally it is important that objectives are clearly set by management and the company when we are talking of fundamental changes in the way of working.
I’m grateful to Margit for her contribution and taking the time to chat. Read more about these topics and more in my featured report in BioProcess International’s May issue: Continuous Processes — Disposables Integrate Upstream and Downstream Processing.