Send to printer »

Feature Articles : Aug 1, 2009 (Vol. 29, No. 14)

Bringing Greater Efficiency to Antibody Manufacturing

Biological Production Challenges Form Basis for Debate at Annual Bioprocessing Forum
  • Susan Aldridge, Ph.D.

At the “Biological Production Forum” held recently in Dusseldorf, there were many lessons for those involved in monoclonal antibody manufacturing. In a wide-ranging presentation, Günter Jagschies, Ph.D., senior director R&D, strategic customer relations, GE Healthcare, spoke about future requirements for monoclonal antibodies.

According to a recent PhRMA (Pharmaceutical Research and Manufacturers of America) report, there were over 190 antibodies in clinical development in 2008. Given that the success rate in Phase I is low, it is unlikely that more than 50 new antibodies will actually come onto the market. But how many of these will require large-scale manufacture?

Dr. Jagschies has reviewed the best-selling antibodies and determined that the annual amounts needed range from 1,197 kg (Remicade) to 25 kg (Vectibix). He believes few future monoclonals will need large-scale manufacture, and that the chance of winning market share from the big five (Avastin, Herceptin, Rituxan, Humira, and Remicade) is small. “But there are some monoclonal antibodies in development for other indications, such as osteoporosis and Alzheimer’s disease,” he added. “If these come through, then they may be candidates for large-scale production.”

Titers of monoclonal antibodies have gone up dramatically in recent years and now average 3–5 g/L, while 10 g/L and even greater are possible. This has implications for the economics of large manufacturing facilities. These higher titers spell overcapacity. “We will see people closing down, mothballing, or selling off their facilities. No one is going to build these dinosaurs of capacity,” said Dr. Jagschies.

Overcoming Overcapacity

In the meantime, there are ways of making facilities work despite overcapacity. Smaller reactors, a smaller plant, high-capacity resins, flow-through systems, and two-step purification can all make a contribution. He also suggested reducing hardware, like tanks, by making buffer from concentrate and doing direct transfer between steps to avoid using storage tanks.

Dr. Jagschies’ analysis shows that higher titers do not necessarily lead to decreased cost of goods (COGs), despite popular belief. Beyond 3–5 g/L, the frequently described strong decline of COGs with increasing titer is relevant only if you actually need to manufacture the amount of drug this enables you to make. “The business message is different from the science message,” he said. Nor, he believes, are there any real economic or technical drivers for alternative production systems for monoclonal antibodies.

In the future, process development teams will focus more on understanding the process rather than on economic optimization through increasing the titer. There will, therefore, be more emphasis on quality by design and, it is hoped, less on QA/QC as a consequence, particularly important if one produces a given quantity with smaller batches, which leads to the release of more batches.

GE Healthcare is concentrating on flexible, smart manufacturing with features that include light, mobile units; segregation rather than dedication; and adding and taking away capacity (rather than rebuilding). Straight-through processing, with in-line adjustment, is seen as important. The firm is also looking at periodic counter current continuous processing, which is a way of getting 50% better utilization of resin in purification, according to some experts. High-throughput process development will also play an increasing role in the biopharma industry, concluded Dr. Jagschies.

There is still a great deal of interest in upstream improvement, however. Robert Young, Ph.D., principal scientist, cell culture process development, Lonza Biologics, described some innovative approaches to improving cell-line productivity. These experiments involved the glutamine synthetase expression vector, which is widely used for monoclonal antibody production.

Use of a stronger promoter, alternate arrangement of transcription units (i.e., heavy chain upstream of the light chain instead of the standard arrangement of light chain upstream of the heavy chain), or insertion of a transcription blocking element between the units did not improve cell-line productivity, pointed out Dr. Young.

The researchers then tried GeneArt’s gene sequence optimization service, which involves altering codons without changing the protein sequence to produce the codons that are optimal for host cell expression. This approach did lead to higher productivities.

“Steps downstream of transcription could be the rate-limiting step in productivity,” surmized Dr. Young. Therefore, looking at transcription itself, as well as translation, assembly, secretion, or a combination of these steps are all important aspects of increasing the concentration of monoclonal antibody in the production medium.

Tackling Productivity Bottlenecks

With collaborators at the University of Kent, Lonza looked at the impact of manipulating the start of translation, focusing upon a eukaryotic initiation factor (eIF4E) that appears to be a good target for cell-line engineering. In house, Lonza is trying to relieve possible productivity bottlenecks by testing alternative signal sequences, which may lead to more efficient secretion.

Dr. Young commented that, although the standard 24-well vector screen is good, it does have limitations because results may not transfer to suspension cultures. There is, therefore, a need for a high-throughput suspension-based vector screen. He concluded that it may be necessary to alter more than one component of a system to optimize productivity

Downstream operations were also well covered at the meeting, with Merck KGaA launching a new smart  cation exchange resin, Eshmuno™ S. The product has been specifically designed for the purification of monoclonal antibodies and offers an alternative to Protein A.

Merck is already known for application of tentacle technology in its Fractogel® resins. In tentacle chemistry, the ligands are more sterically accessible. Thus, the resin has a higher binding capacity for the target biomolecule.

In Eshmuno S, tentacle technology is applied to a new hydrophilic polyvinyl ether base matrix, which allows for much higher flow rates. Experiments show that Eshmuno S has a dynamic binding capacity around 50% higher than other cation exchange resins, according to Merck.

Using Eshmuno S in place of protein A can reduce purification costs by up to 30%, Merck scientists claim. It is also applicable in other purification steps like removal of host cell protein.

Learning from Vaccine Production

Many of the lessons learned in vaccine production could also be applied to monoclonal antibody manufacture, according to  Alfred Luitjens, senior scientist, upstream process development at Crucell. He spoke of how the company is dealing with the need to manufacture a new TB vaccine for the global market because the old BCG vaccine is not effective.

Crucell, in collaboration with the Aeras Global TB Vaccine Foundation, has been developing an improved vaccine, now in Phase II, using recombinant adenoviral technology. The product is being manufactured in its PER.C6® human cell line. 

Crucell knew that a scale-up to 10,000 L  capacity would be needed at the original productivity of the process, in order to meet the demand for 100,000 to 200,000 doses a year. It wanted to fit the process into its  existing facility, however, and chose to focus upon intensifying upstream and downstream processing to boost productivity, rather than merely scaling up.

The result is the iMAP (intensified manufacturing for adenovirus production) program, whose upstream elements include intensified PER.C6 production, large volume high density cell banks, and intensified virus production with adjustment of medium formulation to decrease the volume used.

Comparing standard production with iMAP reveals a decrease in process time from 44 days to 22 days and a decrease in cycle time from 14 days to seven days.  In addition, the process can be carried out in a 1,000 L bioreactor, which, according to Luitjens, brings Crucell close to the goal of a 10–20 fold intensification of its manufacturing process.

Also of note were comments by Alahari Arunakumari, Ph.D., senior director at Medarex, about how increased titers required more attention to be given to downstream clarification processes. Medarex has found that a combination of depth filtration, tangential flow filtration, and centrifugation can be helpful here.

Non-protein A purification with combinations of columns and membranes is also being developed. Meanwhile, Ying Gao, Ph.D., research associate at MedImmune, spoke about how various software tools for COG analysis can help assess the cost of alternative purification strategies and contribute to short-, medium-, and long-term production planning.

Finally, manufacturing in all biopharmaceutical sectors is under increasing pressure to be green, lean, and clean. Flemming Junker, Ph.D., vp biopharmaceutical product supply, Novo Nordisk, spoke about how green manufacturing works at his company.

In 2003, Novo Nordisk set up a climate action program. “Climate change is our business and should be for all companies,” he said. “We need to do something about this, to mitigate the emission of carbon dioxide. It’s good business, because if we are not sustainable, then the price of energy will increase.”

At Novo Nordisk the aim is to decouple economic growth from a rise in carbon emissions. The company joined with the World Wildlife Fund’s Climate Savers Initiative in Denmark and set a goal of decreasing carbon emissions by 10% in 2014 (compared with 2004 figures). “A lot of people said this was not possible,” observed Dr. Junker.

Novo Nordisk approached its goal by looking at how overall production could be made cleaner by doing a deal with the Danish Oil and Natural Gas Organisation (DONG) where energy savings were earmarked for buying green energy and by identifying and implementing energy saving projects with a payback time of five years or less. The company’s efforts have already enabled DONG to build more wind turbines in the North Sea.

According to Dr. Junker, by 2014 all Novo Nordisk’s Danish production sites, including bulk insulin production, will be 100% powered by green electricity. He attested to the importance of small changes such as centralizing WFI production, using condensate for heating, other changes in the cooling systems, and regular leakage tests, in helping to achieve green manufacturing goals.

Responding to a question, Dr. Junker admitted that the environmental aspect of steel versus disposables was a difficult one for the industry. The chair of this session, Prof. Alois Jungbauer, head of the downstream processing group at the Austrian Center of Biopharmaceutical Technology, commented that most of the items used in lab work are disposable these days and these pose an environmental issue that is hardly ever discussed.