Two quarterly meetings of the BioPharma Operations Excellence Consortium were held over the first three months of 2004. On
February 3, the US West Coast chapter of the consortium had its first meeting of the year at Abgenix in Fremont, CA, and on
March 25 in the Netherlands, the European chapter kicked off its first meeting, which was hosted by Centocor Leiden.
ABGENIX MEETING
The Abgenix meeting hosted 40 professionals from over 15 companies and focused on contract manufacturing, a hot topic in an
industry periodically beset by capacity hortages. In addition to the presentations and group discussions, participants were
given a facility tour of Abgenix's large-scale manufacturing plant.
One of the more intriguing sessions of the day was a presentation by Joe Dillon, vice president, supply chain, of Silterra,
a semiconductor contract manufacturing organization (CMO) based in Malaysia. Dillon discussed the industry's background and
how the consortium discussions reminded him of the semiconductor industry 20 years earlier. He also shared that since the
'80s, the semiconductor industry has made major advancements that biopharmaceuticals companies might want to emulate.
As an example, Dillon noted that advanced manufacturing execution systems allow customers of semiconductor CMOs to have full,
real-time visibility of their products in the CMO plants. He also mentioned that there are a large number of "fabless" companies
that focus on product development and outsource all their manufacturing to CMOs, mainly located in Asia. Plus, according to
Dillon, manufacturing technology drives improvement in semiconductors, and, as a result, a semiconductor plant can expect
to turn over 33% of its manufacturing equipment every three years. He added that labor only constitutes 5% of the product
cost, while the tax advantage is the real driver to building CMOs in Asia. The semiconductor industry can thoroughly inspect
each product made, Dillon said. He concluded by noting that although there are many differences between the industries, especially
related to regulatory requirements, there are enough similarities that the future of biopharmaceuticals with respect to contract
manufacturing might look much like the semiconductor industry.
The other presentations of the day were given by Steve Chamow, vice president, process sciences at Abgenix; Karen Walker,
director of quality at Abgenix; Patrick Hanley, Sr., manager of packaging at Genentech; Sid Advant, director of technology
transfer and clinical operations at Diosynth; and Allan Harmon, plant manager for Baxter BioScience.
The main focus of the presentations given by Walker, representing the development side, and Advant, standing for the CMOs,
was the need to develop frequent communication between the organizations as well as a clear, focused set of expectations from
the technology transfer to the development of the quality agreement. Harmon's presentation focused on Baxter BioScience's
transition of its Hayward, CA facility to transition from supporting internal customers to opening its doors to external customers.
For Baxter BioScience, becoming a contract manufacturer required the establishment of a customer-support organization, as
well as improvements to the physical plant to provide added transparency to customers and offer visitors a better understanding
of the current manufacturing processes.
Abgenix's Chamow discussed the challenges of increasing efficiency in biologics production as well as the key factors to increasing
titer. Most of the newer, large-scale production facilities coming on-line are focused on batch production. A discussion developed
about perfusion vs. batch fermentation. From a manufacturing perspective, perfusion production theoretically can be more
efficient, requiring smaller vessels and fewer changeovers compared to batch processing. This is especially true for processes
where large quantities of material are required and suites can be dedicated to products. Many limitations of perfusion were
discussed as well. Current technology and equipment constraints have made large-scale production challenging. Also, perfusion
requires a long-term, up-front investment for validation. Instead of running three batches over two months to validate a batch
process, one may need three runs over at least six months to validate the perfusion process. Because the biopharmaceutical
industry is still young, it is too early to tell which technology will become the standard in the end — but batch processing
has a healthy lead.
