Blog from the 2012 Biomanufacturing Summit On January 31^st at the Biomanufacturing Summit there was a manufacturing technology workshop titled “Analysis of Upstream Cost of Goods” presented by John Bonham-Carter, Vice President, Sales and Business Development for Refine Technology. This was a sponsored presentation where Bonham-Carter presented arguments that perfusion bioreactor systems will be used in the future for biomanufacturing of therapeutic proteins and antibodies and will significantly lower the cost of goods. In a recent blog on the Dish titled “Strategies for Improving Antibody Production in CHO Cells,” perfusion bioreactors were covered as possible way to improve antibody production, in addition to cell culture media supplements and gene synthesis.

The vast majority of biotherapeutics today are manufactured using suspension CHO cells grown in fed-batch culture (typically over a 14-day batch period). Over the last couple decades the science of fed-batch culture, media design, and cell engineering has enabled the yield of product to increase from the ~ 1g of product for each liter of medium to the ~ 6 g/L that is typical for most current production processes. An optimized batch culture system can usually achieve a maximum cell density of 10-15 million cells/ml. John estimated that the use of typical 10,000 L stainless steel bioreactors produce therapeutic antibodies at an average cost of $500 g/L (in 2010) with the cost of facility development and maintenance being a substantial portion of the cost.

John predicted that perfusion systems could potentially lower the total cost of therapeutic antibody production to ~$50/g. The perfusion system that John described is able to reach 70-100 million cells/ml and can maintain healthy CHO cells in culture at that density over an ~3 week period due to the continuous perfusion of fresh culture medium into the bioreactor. Over one year, such a system could produce 10x the product yield compared to a typical stirred tank bioreactor of the same capacity. Alternatively the perfusion system could be used to produce the same yearly yield of antibody in 1/10^th the size of bioreactors. Thus, the use of perfusion should enable the replacement of typical 10,000 L bioreactors with 1000L bioreactors without negatively impacting the yearly yield of manufactured product.

Technically, cost savings derive from several areas. First, large facilities are not needed. 1000L bioreactors can easily be installed in existing facilities. This approach, in turn, can save hundreds of millions of dollars per year typically allotted by large biopharma for facility investment and upkeep.

The other cost savings are theoretical and consist of savings through improved product quality and consistency. In batch culture the composition of the cell culture medium is constantly changing as cells consume the nutrients and excrete waste product. In contrast, the perfusion of fresh medium maintains cells in a constant medium environment. Thus, the use of a perfusion culture system could translate into improved consistency of glycosylation pattern and product consistency compared to batch culture systems.

One drawback of perfusion systems are the large volumes of medium required. John estimated that 1.5-2x reactor volumes of media will be required daily to maintain cells at maximum cell density. Thus, large tanks of media are required which, in turn, negates some cost and space saving potential. However, such systems he argues have the potential to produce 1g of product/day per liter of medium and achieve a theoretical cost of production in the $50/g range.

Interestingly, there is no “out of the box” cell culture medium designed specifically for perfusion systems. Little scientific data is available on the nutritional requirements of cells grown in perfusion vs. the well-established knowledge of cells grown in batch culture. There are likely to be substantial benefits from the use of a perfusion-optimized medium. Hypothetically, a medium designed to produce lower levels of the inhibitory waste products such as lactate and ammonia would translate into a lower requirement for medium replacement. The use of glutamine alternatives (ala-gln dipeptide) or albumin can improve cell metabolic efficiency and reduce the generation of waste products. Glutamine alternatives include Glutagro by Mediatech and Glutamax by Invitrogen. Recombinant albumin can be found through the following suppliers: InVitria’s Cellastim, Sigma’s recombinant albumin, Fisher Scientific’s recombinant albumin, Sheffield Bioscience’s rAlbumin ACF and Mediatech’s cellgro rhAlbumin.

For more information on using recombinant albumin to improve cell metabolic efficiency, please see “Optimizing Overall Manufacturing System Performance through the Use of Animal-free Cell Culture Supplements,” a session at the Biomanufacturing Summit, January 27, 2010. Presented by Matt Croughan, Amgen Bioprocessing Center Keck Graduate Institute, Claremont, CA

What are your thoughts? Are perfusion bioreactors the wave of the future and the best way to improve CHO productivity? Please comment.