This roundtable took place on Tuesday, 5 June 2018, in the BioProcess Theater at the BIO International Convention in Boston, MA. (Left to Right) Moderator: Patricia Seymour, senior consultant at BioProcess Technology Consultants, Inc. Panelists: Jennifer Michaelson (vice president of preclinical research and early development at Cullinan Oncology), Weichang Zhou (senior vice president and chief technical officer at WuXi Biologics), Michael Kaufman, senior vice president, CMC, at Mersana Therapeutics), and John Lee (senior vice president of pharmaceutical development at Decibel Therapeutics)

Moderator Patricia Seymour, with John Lee, Michael Kaufman, Jennifer Michaelson, and Weichang Zhou

Following introductions of the panelists and their companies’ technologies, moderator Patricia Seymour began the discussion about challenges related to choosing different modalities and addressing related manufacturing concerns.

Targeting Modalities
Michaelson began by describing how Cullinan Oncology selects its targets and modalities, how it approaches those early phase decisions, and what its primary driver is to get into the clinic as quickly as possible.

She talked about challenges in oncology, particularly immune-oncology, which is a highly competitive and fast-moving area for several of Cullinan’s projects. Michaelson highlighted the goal of accelerating the timeline for chemistry, manufacturing, and controls (CMC) development, hoping to reduce that to 15 months. Cullinan wants to accelerate progression of multiple candidates in parallel at the early phases of CMC development. In such early stages, additional developability data coming in from orthogonal assays on various molecules can speed the process of narrowing down candidate selection.

Another challenge is to find a contract manufacturing organization (CMO) that provides a “one-stop shop” experience, offering broad coverage of all aspects of CMC development, from DNA to the drug product. Cullinan also is managing many contract research organizations (CROs) for preclinical development. So a goal is to bundle as many activities together as possible under a single CRO. A successful CMC campaign rests on finding a high-quality contract development and manufacturing organization (CDMO) with a proven track record. Such a partner must be flexible and interactive, providing good service and partnership toward molecular development.

Speed, Product Development, and Complex Supply Chains: Seymour agreed on the importance of evaluating different variants early to save time in cell-line development. Speed is a challenge in such a competitive space, especially if a company’s exit strategy is to out-license as well.

Lee talked about the approach that preclinical-stage Decibel takes toward product development. “In addition to the multiple therapeutic modalities that we’re looking at, we have to develop multiple routes of administration,” whether through traditional oral dosages or local injections. He mentioned the importance of building a baseline understanding of how different modalities intersect with different routes of administration.

Because fewer CMOs work in audiology, Decibel invests in specific companies to build those capabilities and enable smooth transition from preclinical discovery into clinical testing. Decibel also tries to leverage therapeutic adjacencies that may be applicable to the products it is developing and to find CMOs and CDMOs with applicable expertise. Examples are products developed for local administration to mucosal surfaces and small-volume ocular products.

Complexities in Antibody–Drug Conjugate Development: Kaufman spoke about the complex manufacturing and the supply chains for antibody–drug conjugates (ADCs), as well as the unusual regulatory environment. Not only are the drug substance and the drug product subject to investigational new drug (IND) application sections and regulations, but the antibody is making INDs “enormously complex.” Manufacturing work has to be staged so that the drug, linker, and antibody all come together to be further processed downstream. Kaufman said that finding available bioreactor capacity is a concern, even though more facilities are being built, but that companies must plan at least a year out for scheduling antibody manufacturing.

For his company, the hardest task is to find external CDMOs with skill sets to supply the linkers. But all three components have to be on manufacturing schedules that run parallel. The ultimate step is to convert all three elements into one ADC, which for his company is outsourced. He noted that “it is a little nerve-wracking when you bring these millions of dollars of components together in the hope that one reaction works really well.”

Zhou added that another difficulty in ADC development and manufacture is that one company might make the antibody and another the toxin and a third the linker. So managing the timing of those shipments can be critical.

Seymour agreed that the risks are high. Partnering with good CDMOs is critical because if any one component goes significantly off schedule, that throws everything else off, and regaining a spot in the manufacturing queue could delay a project by months.

She asked Zhou to comment on how WuXi deals with such complex time lines and technical issues that call for multiple elements to come together amid high technical and regulatory compliance risks. Zhou emphasized the importance of building in capabilities and capacities. In addition to speed is the issue of process robustness. On the antibody side, his company starts by building a good platform cell line and works toward shortening cell-line development timing before using a platform manufacturing process. WuXi’s goal is to reduce the R&D timeline to below 15 months by performing multiple parallel developability studies for multiple candidates at the beginning for developing that initial cell line.

Facilities of the Future
Seymour asked the panelists to offer perspectives about future facilities. What kind of scales will be in demand? What is that going to mean for access to capacity?

Michaelson said that the challenge is knowing what capacity will be needed so that a company doesn’t have to wait to take its molecule into development. But she also noted that in oncology development, the biggest challenge is not CMC. “Right now it’s how all the molecules in development will find their right use for the right patient in the right combinations because there are so many molecules currently in development and so many combinations being tested. It’s probably going to take time in the clinic and some carefully planned translational work to find the appropriate molecules for the appropriate patients and indications.” She thinks that CMC becomes secondary to trying to predict scale, but she warns against underestimating how successful some immunotherapy products might be — and how quickly those could be developed and moved into commercialization.

Personalized Medicines and Capacity: Seymour asked about the impact of personalized medicine on manufacturing. “The scales could become quite small, and they could become quite mobile, depending on what technologies or modalities you’re talking about. That’s quite challenging, particularly from a CMC perspective. Can you actually set up a mobile facility close to a patient population?”

Lee agreed that the shift from very large manufacturing to more personalized cell- and gene-therapy medicines will necessitate manufacturing flexibility. “Capacity is not going to be the driver; it will be the quality of the material. Lots will be smaller and based purely on drug need.” He said that beyond assessing quality aspects when selecting a CDMO is the importance of understanding that company’s philosophy. Does its business model intersect with a client’s needs?

Kaufman returned to CMC challenges in the oncology space, where developers typically toward the “maximum tolerated dose.” That might not be known until late in a program. Oncology drugs also are tested for different tumor types and cancer indications. So it also can be late in the game before a company knows whether it will need 200,000 or 10,000 doses per year. Finally, if a company gets good results, many breakthrough-therapy and other accelerated-approval schemes become available. All those elements make it challenging to know what the manufacturing scale and timelines will be. One advantage for ADCs is their potency, which can help reduce their supply-chain complexity.

Zhou described WuXi’s approach to scaling out rather than up. The company bases its manufacturing on single-use bioreactors that scale from 200 L to 4,000 L. Because product quality is key, he suggests staying at one scale (2,000 L, for example) in phases 1 and 2. Then by adding more bioreactors instead of scaling up to 20,000 L, a company can lessen risks economically without compromising product quality comparability. Essentially that keeps cell culture at the same scale but satisfies product demand. A company doesn’t have to commit to investment in a large facility early on, and with single-use bioreactors, can build a facility in as little as two years from green field to operation. That allows time for defining a CMC process as well as determining product demand.

With such a strategy, WuXi can build a facility where a product will be needed. That provides supply-chain robustness and mitigates the risk of facility-to-facility validation because the scale does not change while a manufacturer meets product demand in a different geographic location.

Watch Online
Watch the full interviews online at www. bioprocessintl.com/BIO-Theater-2018.