"By 2016, six out of ten of the blockbusters will be monoclonal-antibody-based therapeutics. With patent protection ending on many of these blockbusters by 2020 and this market estimated to be worth $100 billion a year, it is clear why developing biosimilars is attractive,” stated Stuart Melville, purification team leader at Eden Biodesign.
Speaking at the recent “Future Technologies in Downstream Processing” conference at the Stevenage Bioscience Catalyst, Melville added, “Biosimilars also have a shorter development cycle. To develop an innovative biologic, it costs in the region of $800 million and can take around 10 years, whereas for biosimilars it costs around $100–150 million and takes approximately eight years because you can bypass the Phase II dosing range studies.
“However, it is a very competitive marketplace, and Teva and Lonza have already halted their rituximab biosimilar development, so looking at areas such as cost reductions in manufacturing and ways to differentiate your biosimilar other than on price is imperative.”
The difficulty with manufacturing a mAb therapeutic is that the molecule has huge complexity with over 100 million variations of the IgG, and finding the most effective clone that can generate antibodies cost-effectively is challenging.
To reduce manufacturing costs, during upstream processing the focus is on optimizing clone selection and bioprocess conditions, but there is also a need to look at how to maximize antibody yield from the clarification, purification, and polishing steps in the downstream processing phase of production.
New technologies are being used in the initial stage of the downstream process to help overcome process bottlenecks and drive down cost-of-goods (COG) of mAb therapeutics.
Aloke Dey-Chowdhury, senior technical specialist at Pall Life Sciences, discussed the Cadence™ single pass tangential flow filtration (TFF) system, which showed that by using a Cadence module with a flow path equivalent to six TFF cassettes in series, 100 L of bioprocess liquid containing a mAb could be concentrated in-line, reducing the overall step processing time from 18 to 11 hours.
“Eliminating the requirement for recirculation associated with conventional batch TFF allows for continuous in-line concentration and is how we can achieve six hours process time savings,” Dey-Chowdhury explained. “With single pass TFF the concentration process is a function of the flow path and not of time, and so depends on how many cassettes are in the flow path.
“Integrating this TFF technology helps overcome the potential bottleneck in capture purification, but it won’t replace conventional ultrafiltration for all processes. It is a production improvement technology, where increasing mAb yield and product quality are the benefits, which may offset the cost of buying additional TFF cassettes.”
Steve Loftus, Ph.D., technical project leader at Fujifilm Diosynth Biotechnologies, believes that for some stages of downstream processing, using disposables reduces costs. “We decided to trial disposables for some areas of our downstream processing because it reduces plant time and operational expenses. Since we are a multiproduct facility, we need fast changeover between batches. When we’re manufacturing high potency products, using disposables eliminates carryover, reduces cleaning requirements, and improves operator safety.”
Dr. Loftus showed an example where Fujifilm Diosynth had trialed its existing lenticular filtration system in a clarification stage against a Sartoclear® (Sartorius Stedim Biotech) disposable depth filter.
“With the lenticular filter we had to load cartridges into a bell housing and this was difficult to do. Using the Sartoclear filter is easier as there is no cleaning and less hold-up volume, which means there is no buffer spillage when we remove the disposable filter. Using this filter, we estimate that we save on operational costs and 12 hours of plant time.”
Capturing the Antibody—New Tricks?
The capture step can be the most expensive one in downstream processing due to the use of costly Protein A affinity media. “We don’t want to reinvent chromatography but do want to introduce innovation to improve the process and decrease cost of goods, handling time, and footprint of the process,” stated Fabien Rousset, Ph.D., materials/project manager of Novasep.
He discussed AbSolute® High Cap, a new type of Protein-A-based chromatography media, which uses beads with a more rigid structure made of silica and coated with epoxy resin. Dr. Rousset showed examples of how the structure of the bead enables increased flow rates of up to 600–800 cm per hour.
By increasing dynamic binding capacity and utilizing smaller columns with less volume of Protein A resin, users can employ higher flow rates and potentially double their productivity, according to Dr. Rousset. At the 100 L scale, he estimated that this increase in productivity and reduction of column sizes could save as much as €2 million ($2.6 million) of operational costs by reducing the amount of buffer, resin, column packing, and associated labor.
Dr. Rousset also presented another technology called sequential multicolumn chromatography using the Prochrom® BioSc® system, which automates the loading and regeneration of a series of capture chromatography columns in sequence using continuous UV and pH sensors to monitor column loading.
“Classically, if you load a column at 100 cm/hr flow rate the column is 83% loaded, at 200 cm/hr it is 68% loaded, and at 400 cm/hr it is 38% loaded before leakage of the protein occurs (dynamic binding capacity),” he explained. “By loading and reloading a series of columns in sequence instead, you can replace batch mode and utilize 100% of the media capacity (static binding capacity).”
Membrane Absorbers for Polishing
Again disposable technology is increasing in popularity for the final polishing step. “The market applications for membrane absorbers have grown by around 27% in the past six years, and they are being used in places instead of resin chromatography to remove contaminants and viruses. They are mostly used in the final polishing step,” commented Juan Pablo Acosta Martinez, Ph.D., project manager at Sartorius Stedim Biotech.
“The main reason for this rise in use is that membrane absorbers can increase flow rate by up to 30 times. Also the bed heights are smaller, only 4–8 mm, and so take up less plant space. Being disposable means there is no cleaning or cleaning validation time required either.”
To show the effectiveness of membrane absorbers in the polishing step, Dr. Loftus presented an example where his group trialed a traditional anion exchange flow through column against a disposable Sartobind® membrane.
“On average it takes us about 48 hours to clean equipment, pack a column, and equilibrate it for use. We estimate that using the membrane absorber instead of traditional packed-bed chromatography, we save on operational costs, as well as reduce plant time by 95%, from 64 hours to just three.”
Everyone at the event agreed that there is a significant time and cost saving when using disposables in all stages of downstream processing of mAbs because they increase process efficiency. Many delegates also cited the need for greater product development in this area.
“Disposable cell disruption systems are not available, and by using 50 L to 500 L disposable chromatography columns we are at the physical capability limit of plastic at the moment,” explained Andrew Clutterbuck, biomanufacturing engineer at Merck Millipore.
“Disposable columns cannot be operated beyond 3 bar pressure, so we have to use steel for the 1,000 L plus purifications. However, perhaps automated continuous chromatography methods are the answer. Or as antibody titers are increasing, we may not need larger columns, and with improvements in upstream processing, in the future we might be able to use the existing disposable chromatography technologies.
“There is an increasing need for development of, as well as early integration of, disposables and automation in downstream processing. This will drive future success and make production of biosimilars more cost-effective,” Clutterbuck concluded.