Developments in Continuous Cell Culture

Significant benefits, technologic hurdles

Continuous cell culture is part of a trend in bioprocessing towards highly productive, streamlined, and interconnected unit operations. While continuous processing has a long history in various industries, from foods to consumer items to high-tech electronics, regulatory concerns -- some real, some self-inflicted -- have stalled widespread adoption of continuous bioprocessing. That is beginning to change as the industry, particularly biotech, acclimates to regulators’ desire for risk- and science-based production decisions.

As far back as 2000 Dr. Andrzej I. Stankiewicz of Delft University in the Netherlands, writing in Chemical Engineering Progress, identified continuous cell culture as a type of process intensification. Normally that term describes processes in which operations are combined, eliminated, or otherwise streamlined, a definition that Stankiewicz felt was too limiting: "Clearly, a dramatic increase in the production capacity within a given equipment volume, a step decrease in energy consumption per ton of product, or even a marked cut in wastes or byproducts formation also qualify as process intensification,” he wrote.

Stankiewicz’s thinking has caught on in mainstream bioprocessing. Michael Phillips, Ph.D., Director of Next-Generation Process R&D at MilliporeSigma (Darmstadt, Germany), positions continuous cell culture squarely within of process intensification. “But adoption has been slowed by regulatory and technical issues,” Phillips says.

Cell retention is key

Batch and fed-batch are the most common type of mammalian cell culture, regardless of scale. In batch cultures, cells, media, and nutrients are locked into a bioreactor for up to three weeks with nothing added or removed. Fed-batch culturing involves adding nutrients at varying intervals in the hope of extending the culture and improving yield. In both cases product is harvested at the end, in one batch.

In continuous cultures product is perfused constantly or at regular intervals, fresh media and feed are constantly added, and in many cases cells are bled off to maintain an optimized population of healthy, productive cells. Perfusion cultures produce prodigious quantities of product for a given volume, over several months, compared with batch culturing.

While several commercial bioprocesses employ perfusion cell culture or some aspect of continuous purification (downstream processing), the goal of end-to-end continuous operations remains elusive for approved biomanufacturing processes.

Due to their continuity and productivity perfusion cultures only make sense for large-scale processes, where downstream purification has been designed specifically to accommodate continuous production, or in scale-down or developmental models of those processes.

Downstream operations may in fact be the determining factor in deciding on continuous or batch cell culture. “Purification must run for the same duration as the perfusion culture, and more or less continuously,” notes Peter Levison, Senior Marketing Director for Downstream Processing at Pall Life Sciences (Portsmouth, UK). “With batch processing clarification and individual chromatography steps might last one day or less.” In continuous processing, which goes on for weeks, manufacturers must decide on what to do with accumulating product. Storing product until sufficient quantities build up for purification in batch mode is one approach. In an ideal (but as yet unrealized) scenario product would feed continuously into a purification train that operates for the duration of the culture.

Continuous cell culture is getting smaller...

Like all unit operations the optimization of continuous cell culture depends on the ability to scale up and scale down -- and the level of trust developers can place in various small-scale models. Yet the decision to go continuous is usually based on data from batch processes.

The earlier in development a team adopts continuous cell culture the greater the potential for full optimization. “Working in perfusion mode from the smallest practical scale pays dividends in understanding the technical parameters -- what needs to be controlled and what doesn’t,” says Ken Clapp, Sr. Manager for Applications Technology Integration at GE Healthcare (Marlborough, MA).

Hence the increased desirability for cell retention devices that properly size for the benchtop, that pair well with glass, stainless steel, or single-use bioreactors, and that deliver process understanding at the scale at which process developers are used to working. “The development stage will proceed more smoothly, quickly, and at less cost than if you wait until you’re up to fifty liters. You’ll have more information on media feed, cell bleed, cross-device pressures, and cell viability. You get your answers then and there,” Clapp adds.

As with batch cultures yields become secondary to process issues during continuous cell culture development. Engineers must consider the right size of the retention device, the type and manner of recirculation and cell removal, how to establish secure connections between culture and harvest, and media/feed replenishment. All with the goal, says Clapp, of envisioning how the process will look at whatever its terminal scale. 

Given the constant medium replenishment demanded by continuous cultures, producing and storing the nutrient feed diminishes the space and footprint advantages of continuous cell culture. Processors must weigh the advantages of making up one large batch, or scheduling the constitution of many smaller batches.

...and even smaller

Perfusion cell culture is usually moot outside of manufacturing-scale bioprocesses, or in experimental models for which manufacturing is a goal. Developers looking for a few grams of material for characterization or preclinical studies need look no further than batch processing.

Most cell retention systems sold today operate at bench scale and above but very few options exist for below about two or three liter volumes, which has become the de facto preferred scale of process development cultures. Yet developers would readily adopt smaller retention devices, operating at lower volumes, to reduce development costs related to cells, media, and feeds even further.

Spin filters, cell retention devices that use centripetal force to expel proteins through a membrane while retaining cells, do in fact operate at sub-liter scale. Spin filters have been around for years, vendors still sell them, and the devices remain popular in research laboratories. The drawbacks of spin filters outweigh the benefits for companies aspiring to large-scale continuous cell culture, however.

John Poppleton, Marketing Director at Applikon Biotechnology (Foster City, California), explains: “Since spin filters are not available at very large scale, extrapolating results to beyond experimental scale is problematic. It also takes quite a bit of work to optimize them, and spin filters tend to clog after one to three weeks of use. This eliminates most of the advantages of long-term perfusion cultures.”

As production scales rose and biotech came into its own, during the early 1990s, tangential flow filtration cartridges, and later hollow fiber systems, became the cell retention devices of choice. “Some real breakthroughs took place back then in terms of practical understanding, and pairing those devices with smaller bioreactors,” Clapp says. “Internal and external filters worked quite well at various scales, with the external spin filters having the advantage of being replaceable during a run, using steam bleed/block valves. This overcame one of the main problems with filter-type methods: clogging.”

Unlike conventional membrane-based perfusion devices Applikon’s Biosep cell retention system is a non-fouling filter that relies on acoustics to separate trapped cells from dissolved protein products. Biosep allows continuous harvest from the perfusion process, while the retention device keeps cells inside the bioreactor at an ideal concentration of about 100 million cells per milliliter.

Biosep serves a niche that has been difficult to fill, namely very small-scale perfusion cell culture that has become the mainstay of both scale-up and scale-down studies. In addition to serving production perfusion scales of 1000 liters per day and pilot scale (200 L/day), Biosep covers the low end of development scale, down to one liter per day of perfusate cell-free, product-containing media.

Monitoring and control of continuous cell cultures

Automated process control is highly desirable in all commercial cell cultures, particularly for dynamically adaptive continuous processes. Without constant measurement of pH, cell density, dissolved oxygen, etc., and automated control based on those parameters, the optimization of nutrient feeds, cell removal, and harvest rate in perfusion culture could never occur.

That is why Applikon has incorporated the Futura capacitance sensor from Aber Instruments (Aberystwyth, UK) into its Biosep cell retention device. Futura measures viable cell mass inline, thus enabling automation of perfusion cell culture. Specifically, Futura is used to control feed and harvest rates.

Many sensors, in use or under development for continuous cultures, had their beginnings in batch mode. Stratophase (Hampshire, UK) has traditionally specialized in a closed-loop, real-time, non-sampling system for controlling the addition of nutrient feed to fed-batch processes through its Ranger® product. Ranger supplies nutrients on demand according to metabolic need.

Stratophase CEO Simon Saxby notes that unlike conventional sensors and controllers Ranger relies not on conventional measurements of pH, dissolved oxygen, cell density, etc. but on the rate of change of the medium’s relative refractive index as a surrogate for culture health. In doing so it adds a degree of continuity to what are typically non-continuous cell culture processes.

Perfusion cultures present a more complex scenario in that the critical operations that keep it going involve harvest (which occurs automatically through the cell retention device), cell bleed, and media/feed replenishment.

“Operators of continuous processes tend to rely on offline sampling to determine feed rate over the next time period,” Saxby says. “That determination, say to add one liter of feed per hour, does not take any real-time changes in metabolism into account. In that regard it’s not much different from fed-batch processes. Having adaptive feed control should significantly improve the output of continuous processing, and allow cells to be maintained for longer, in a healthier, more productive state.

From the Ranger system perspective, maintaining a steady state through on-demand feeding should be no different in continuous cultures than for batch. “The difference arises in where to locate the sensor in a perfusion process.”

One additional benefit of Ranger is that it is agnostic as to reaction vessel design, size, or materials of construction. This is an advantage for scale-up and scale-down of continuous processes, which depending on the adopted technology is not always reliably predictable or even possible.

Angelo DePalma is a freelance writer living in Newton, New Jersey, USA.