Replacing Steel and Glass with Plastic

Disposable bioprocessing has overtaken stainless for small production batches of up to several hundred liters.

Single-use equipment entered biomanufacturing as simple bags for holding media and buffers. As mixing technology improved, media and buffer preparation inside the bags became common. By the early 2000s, bags were common for volumes of up to about 200 L for cell culture processes as well, mostly for pilot- and early clinical-scale manufacturing.

Disposable bioprocess containers’ current size limitation of about 2,000 liters would have been more of a barrier 15 years ago than today. Since the early 2000s, when experts predicted an impending biomanufacturing capacity crunch, titers for therapeutic proteins have risen from less than 1 g/L to an average of 2.5 g/L, with approximately 12 percent of commercial processes boasting greater than 6 g/L and higher. These data, from consultant Eric Langer of BioPlan Associates (Rockville, MD), include titers for all protein processes. Figures for monoclonal antibodies, which make up the highest volume and revenue components of therapeutic biotechnology, are undoubtedly higher due to the agents’ familiarity with and adaptability to platform processes.

Why Plastic?

Disposable cell culture containers provide advantages that only make sense in the world of high-cost, high-value biomanufacturing.

Plastic vessels eliminate lengthy, expensive cleaning and cleaning validation of stainless steel bioreactors, which processors must demonstrate after each production run. Single-use process bags arrive at the point of use gamma-sterilized; when the batch is completed, the bags are discarded. Single use practically eliminates contamination and cross-contamination.

Related Article: How to Avoid Contamination in a Microbiology Lab

On relative process economics for stainless steel vs. plastic, the consensus favors single use. Dr. Dethardt Müller, VP Technology Development, Rentschler Biotechnologie (Laupheim, Germany) has calculated that a 1,000 L single-use antibody process costs approximately 12 percent less to run than the equivalent process in stainless steel. Principal savings are in depreciation and maintenance (e.g., cleaning, cleaning validation).

And for flexibility and agility, particularly in multiproduct facilities, single use leaves stainless tanks in the dust. Operators can set up the next batch in hours instead of days. Without the need for steam-in-place sanitization, the bag can go anywhere regardless of the configuration of utilities, and two different products may run in the same suite.

Upstream, Downstream Divergence

Single use is well-established with bioreactors, cell harvest, polishing membranes, and final bulk filtration, but capacities vary depending on culture volume and titer.

For example, the largest single-use bioreactors are 2,000 L working volume. But when cell densities are higher than 20–30 million cells/mL, this quantity of process fluid will overwhelm single-use depth filters for cell harvest or clarification.

“Single use has lagged in purification, particularly for ultrafiltration/diafiltration, polishing and of course for chromatography resins,” notes Parrish Galliher, chief technology officer for Upstream Processing GE Healthcare’s Life Sciences Business (Marlborough, MA).

“Prepacked resin columns are limited in scale and are available up to 30–40 L, but they are still too expensive be single-use and are limited in capacity and only able to handle 2,000 L harvests up to 3–4 g/L titer.”

Polishing membranes for clearing contaminants such as DNA, host cell proteins and viruses are available as single-use cartridges but are limited in capacity for single- use 2,000 L harvests at protein titers beyond 3–4 g/L titer. Similarly, ultrafiltration for product concentration and diafiltration for buffer exchange are available in single-use format, but are limited in capacity beyond 2,000 L harvests at 3–4 g/L. Single-use final bulk sterile filtration cartridges can generally handle high loads from 2,000 L harvests.

Chromatography - A Special Case

Single-use chromatography is simply not feasible due to high resin costs. With a several-million-dollar investment in just the protein A capture column—and a typical mAb process uses two other chromatography steps— bioprocessors and vendors focus more on resin cycle life and resistance to strongly basic cleaning agents than on developing disposable resins. An article or presentation occasionally makes the case for reusing chromatography resins throughout a campaign for one product, then discarding, but manufacturers generally prefer to hold onto their $10,000-per-liter protein A resins.

Downstream processing is where appreciation for disposability slows significantly. A standard antibody purification involves harvest (depth filters or centrifuges), capture (protein A chromatography), cation exchange chromatography, virus filtration, anion exchange, ultrafiltration, and filling with buffer prep/hold and product hold thrown in for good measure.

Related Article: How Downstream Plasma Cleaning Works

The final column, anion exchange chromatography to remove residual impurities, has to some degree been replaced by anion exchange membrane adsorbers, which resemble filters more than columns. This is possible because impurities at this stage have already been reduced to trace levels.

Single-use replacements for bind-and-elute methods like protein A capture and cationic exchange have proved elusive because in these steps the column must have the capacity to bind all of the product as impurities flow through.

The closest approximation to single-use chromatography today is continuous chromatography, which provides disposable-like economics by pushing resins above and beyond normal usability.

Continuous Chromatography to the Rescue?

One such technique under development at ChromaTan (State College, PA) is Continuous Countercurrent Tangential Chromatography™ (CCTC), which is essentially column-free purification.

Instead of being packed in a column, resin flows as a slurry through a series of static mixers and hollow fiber membranes. As the slurry moves through the system it cycles through binding, washing, elution, stripping, and equilibration steps, returning back to the binding step for the next cycle. The buffers in CCTC flow counter-currently to the flow of the resin slurry in each step, extracting either impurities or product. According to president and founder Oleg Shinkazh, this approach conserves buffer and utilizes resin to its fullest. In published studies, CCTC is reported to achieve protein purities of 99 percent with recoveries of 94 percent. Productivity of up to 75 g of protein per liter of resin per hour is as much as tenfold above that of a conventional packed column.

“Our system runs four cycles per hour, versus two to three cycles per shift for conventional column chromatography,” Shinkazh says. “We approach the economics of single use because CCTC uses approximately onetenth as much resin as a column for equivalent results.” The bulk of the resin is therefore replaced by a relatively inexpensive single-use flow path, enabling savings of anywhere from 30–50 percent depending on the application.

ChromaTan is currently performing commercial evaluations with multiple large biotech companies, and has jointly published data with Fujifilm Diosynth and Regeneron. The company has closely collaborated with Professor Andrew Zydney at the Pennsylvania State University Department of Chemical Engineering, and has also entered a manufacturing agreement with Thermo Fisher for high-quality single-use CCTC system components.

Sensors: Critical and Enabling

Based on FDA’s Process Analytic Technology (PAT) and Quality by Design (QbD) initiatives, bioprocessors are expected to achieve process understanding, bring their process under control and generate quality product through deliberate application of real- or near-real-time analytics. The previous paradigm involved “testing in” quality after the fact. PAT and, therefore, QbD are more difficult to achieve in plastic because the sensors and analytic products are not as robust or reliable as reusable probes fitted into stainless steel tanks.

The paradigm shift of moving from electrochemical to optical sensors has singlehandedly boosted the availability and adoption of sensors for disposable bioprocessing, according to Professor Govind Rao of the University of Maryland, Baltimore County. “We can now make sensors that are disposable and can be more readily integrated into process equipment. It’s also opened up the retrofit market because companies will not abandon their glass and stainless steel investment. It becomes a question of adapting optical sensors to the standard parts that these bioreactors are equipped with.”

The downside is that optical sensors are typically not ready for use out of the box. Users are typically tied to a single vendor with their own unique chemistry and architecture. “My vision is one where sensors and disposable systems are designed to be a plug and play like the USB standard used in computers, so that any vendor’s sensor could plug into any other instrumentation or hardware, and would be calibration free,” Rao says. “This is where the industry needs leadership, perhaps a consortium approach to arrive at standards.”

Adoption Barriers

Where plastic is ubiquitous at laboratory scale and during process development, its acceptance declines as a process moves closer to patients. Leading biopharmaceutical companies have significant investments in stainless steel equipment and related infrastructure. The benefit to them from rising protein titers is the ability to run half as many manufacturing campaigns, and thereby half the number of clean/steam-in-place sterilization cycles.

Concerns over leachable and extractable plastic components are valid. Where the human gastrointestinal tract is remarkably resilient and adaptive, every impurity molecule in an injectable or infused biomolecule ends up in the bloodstream. Once there, unfathomably low concentrations may cause immune reactions. Although leachables have not yet caused a catastrophe, the biopharmaceutical has convened several expert panels to deal with them.

Supply and interoperability are related barriers to adoption. Since components of single-use systems from supplier A usually will not integrate with components from supplier B, end users are joined with suppliers in ways they are not, say, with vendors of buffer salts. Redundant supply chain strategies are therefore a must for any biomanufacturers contemplating disposables.

Finally, microbial fermentation, still a significant part of bioprocessing for non-mAb therapeutic proteins, is an area where single-use technology has lagged for years. The mechanics and physics of fermentations are much more strenuous than for cell culture, and processes are much higher volume, hence the low appeal of plastic and its 2,000 L working volume limit.