Ahoy there! How much real gold is in ye olde pouch of clinking coins? Once upon a time, a banker with a set of scales and standard measures could savvy you. Nowadays, cell and gene therapy developers answer to a different calling: to seek advanced treatments that prove safe and effective for unmet medical needs. Yet some things never change. An empirical sense of balance helps you avoid costly setbacks, such as the insidious kinds that could creep in from the cell bank’s inception.

To help teams navigate the lowest risk route to prepare manufactured cellular starting materials for tomorrow’s regenerative medicine doses, RoosterBio’s Joseph Candiello, PhD (Assoc. Director of Product Management) and Joseph Takacs, MS, (Sr. Bioprocessing Application Scientist) recently presented a webinar, “Cell Bank & Scale Up Strategies for MSC & Exosome Manufacturing.” (Watch it HERE).

In brief, this two-part presentation explored why many promising programs can stall before Phase I, how smart choices in cell banking and manufacturing strategy can prevent those delays, and what practical steps developers can take to reach the clinic faster with scalable, phase‑appropriate processes. Joe Candiello (“Joe C.”) opened with key decisions in strategic cell bank manufacturing while Joe Takacs (“Joe T.”) followed with real‑world modeling to demonstrate how effective scale-up strategies for MSC manufacturing can set up a program for long term success in trials and beyond. There was also an enlightening bit of Question & Answer buzz at the end. Interesting? Read on! Or click on the webinar link.

Banking the Cell Product’s “Origin Story”

Joe C. explained that the industry standard for cell banking is a two‑tiered system designed to ensure reproducibility from research through commercial manufacturing. First, a master cell bank (MCB) is created from a starting tissue a cell line and fully characterized to meet quality and safety standards. The MCB serves as the reference source for future banks. In turn, working cell banks (WCBs) are derived from the MCB and used to produce batches for development, clinical, or commercial use. Before advancing, it’s good to become educated on variables that impact MCB and WCB design. Examples of these include whether they will be from autologous or allogeneic sources, what kinds of quality and testing needs they will need, how to manufacture and scale up, and the additional engineering the cells will need.

After sufficient backgrounding, the next step is to design the banks. With practical knowledge of their fundamental requirements to dovetail their intended use, it will be crucial to water test them with clear answers to these questions:

• Will it last through a product or therapies lifecycle?
• Does it meet Quality and Regulatory requirements?
• Does it have geographical quality and logistical flexibility?
• Will it meet future scalable manufacturing needs?
• Is functional variability tested and within CQA tolerances?
• Does it fit into a feasible cost structure for early clinical programs through commercial success?

Joe C next outlined how building a GMP‑grade cell bank can be a complex, multi‑track process that typically takes 12 to 24 months and costs $1.5 to $3 million. Parallel efforts include donor tissue sourcing and testing, process development for both the master and working cell banks, analytical method development for release and potency assays, and long‑term stability programs. Each stage involves tech transfer, engineering runs, and regulatory preparation, all under a quality system. Yet, even in the best‑case scenario, this effort pulls focus and resources away from other critical priorities such as drug product development, understanding CQAs and analytical needs, generation of preclinical data, and regulatory submission preparation.

Given that Murphy’s Law often rears its head out of complex activities, Joe C, outlined several common mistakes to be wary of. These errors can undermine a cell bank’s long‑term utility and force costly delays or rebuilding efforts:

• Building too small of an MCB, which can quickly exhaust supply and require a new bank with comparability and regulatory re‑approval.
• Mismatched MCB and WCB volumes, leading to working banks that are unusable at scale and create long‑term production inefficiencies.
• Inadequate donor screening or cell bank characterization, which can delay regulatory approval, require expensive confirmatory testing, or even invalidate a bank.
• Donor variability affecting potency or manufacturability, resulting in missed lot size requirements and an invalidated cost model.
• High failure rates due to inexperience, which can cause up to two‑thirds of manufactured lots failing from contamination, fill‑finish errors, or user mistakes that force a restart from an earlier milestone.

Joe C. then explained two ways to reduce risk and speed progress to Phase 1. The first is to proactively work with experienced partners who can help avoid common pitfalls in cell banking and manufacturing. Option 1 (Building a Bank) results in a Year 1 investment of ~$1M dollars and a program nearing working cell bank production, just beginning donor evaluation to see if it can meet the program’s needs. Option 2 acquires only what is needed for early clinical milestones by use of pre-stocked, GMP‑ready working cell banks. The latter empowers developers to start regulatory filings and dose manufacturing sooner while postponing larger investments until the program advances.

Figure 1, above, a schematic to illustrate how RoosterBio’s GMP CliniControl™ cell vials, master cell banks (MCBs), and working cell banks (WCBs) can save time, reduce cost, and lower risk for cell therapy product developers en route to human clinical trials.

RoosterBio’s GMP working cell banks fit squarely within the second strategy. By providing pre‑qualified, regulatory‑ready WCBs supported by USFDA Master Files, RoosterBio enables developers to bypass the lengthy process of building a bank from scratch. This allows first-in-human programs to begin donor evaluation, regulatory filings, and Phase I dose manufacturing much sooner while economizing resources. Because these banks are produced using standardized, scalable processes, they also serve as a strong foundation for later-Phase creation of one or more program‑specific MCBs and for seamless transition to larger‑scale manufacturing as the therapy advances.

Incidentally, choosing the right bioprocess partner who uses the optimized toolset of GMP media and cells matters, too. Joe C. closed his section of the webinar by showing a cool video of MSC (mesenchymal stem/stromal cell) expansion that compared two different cell+media pairings. [1] Can you guess which system resulted in 63% less media volume, 1/3 the growth time, 2-fold increase in net cell yield with a 40-to-55% reduction in total cost?

When a Plan Comes Together (for scalable manufacturing)

Most everyone loves it when a plan comes together, but what goes into a solid plan for cell therapy scale up?  Joe Takacs provided the important “deets” through the next half of the webinar. He first emphasized that each step of cellular bioprocess development builds on the last, and that early platform decisions have lasting consequences for cost, timelines, and comparability as programs advance.

The high-level, end-to-end segments of the operation are:

• Develop a Cell Bank – Establishes the foundational source material for all future manufacturing.
• Develop 2D Process and Establish CQAs – Builds understanding of how the cells behave and what critical quality attributes must be maintained.
• Develop a Scalable Process – Ensures that the manufacturing approach can grow to meet increasing demand without losing product quality.
• Transfer Process to GMP Manufacturing – Finalizes production under regulatory standards so material can be used in clinical trials.

And yet, as an essential prerequisite to the above, Joe T. explained that effective scale‑up begins with a good plan. This means one must effectively

1. Define the future, what it looks like and work backward to plan
2. Build a model with conservative assumptions that estimates a range of needs
3. Use the model to project manufacturing scales required through clinical progression; and
4. Develop and implement a scalable manufacturing platform to support each clinical trial phase.

By estimating annual dose requirements and accounting for overfill, downstream processing losses, and testing reserves, developers can calculate the number of lots and doses needed per run.  Joe T. accordingly presented an example model to show how lot size requirements informs platform choice. In this example, a future commercial-stage therapy is assumed to eventually need 10,000 doses per year, each containing 100 million viable MSCs, which equates to over 1 trillion cells annually. To meet this ultimate demand, Joe worked backward to calculate the necessary run rate. He estimated 25 lots per year (one lot every two weeks) and applied a 10 percent scrap rate, leaving 23 successful lots annually. Each lot would need to yield about 435 doses, plus 100 extra doses for release and retain testing, bringing the target to 535 doses per lot.

After adjusting for 30 percent overfill and 30 percent downstream processing loss, the per‑lot production target rises to roughly 390 billion viable MSCs, resulting in a realistic annual target of more than 2 trillion cells. These calculations would then serve as the basis for comparing the suitability and efficiency of different 2D and bioreactor platforms. At this scale of production, 2D cell expansion systems aren’t practical. However, bioreactor manufacturing platforms can practically overcome many limitations of traditional methods by significantly reducing labor requirements, including fewer operators per run. 200-2000L “3D” systems already under evaluation may prove to be an adequate fit.

Figure 2, above, showing plausible baseline assumptions to undergird de novo costing models to fit different scales of production required for advancement of an MSC cell therapy, from Phase I to Phase III clinical trials.

Working backward from a commercial product, Takacs then provided cost models for its accompanying Phase I, II, and III trial needs. He assumed flat GMP labor costs of $50,000 per operational day, average material and consumable costs specific to each platform, and $150,000 per manufactured lot for release testing. The model also included a 10 percent retain requirement per lot, an average yield of 70,000 cells/cm² for 2D platforms and 700,000 cells/mL for bioreactors based on RoosterBio’s internal data, and adjustments for 30 percent downstream processing loss and 30 percent overfill. These standardized assumptions allowed him to compare the efficiency, scalability, and total cost of different manufacturing approaches at each clinical phase.

For Phase I, Joe Takacs modeled the cost of producing more than 9.3 billion viable MSCs to treat 25 patients with two doses each. His analysis showed that both a 40‑cell stack process and a 35 L bioreactor could meet the target in a single run for less than $1 million in GMP manufacturing costs. These scales provide a practical entry point for early clinical production while using platforms that can later support larger‑scale manufacturing as the program advances.

For Phase II, Joe T. projected the need for more than 19 billion viable MSCs to treat 50 patients with two doses each. The modeling showed that either a 60‑cell stack process or a 35 L bioreactor could meet the target with just one production run while keeping GMP manufacturing costs below $1 million. These options balance efficiency and cost while maintaining continuity with platforms that can scale further for later phases.

For Phase III, Takacs estimated a need for about 57 billion viable MSCs to treat 150 patients with two doses each. His modeling showed that two runs using a 50 L bioreactor could meet this requirement at a lower total cost than either smaller‑scale bioreactors or traditional cell stack systems. While a 200 L bioreactor could also achieve the target with fewer runs, the higher upfront investment and qualification requirements make the 50 L platform a more practical choice at this stage, while still aligning with future commercial scale‑up.

Joe Takacs accordingly closed his section with these words:

“It is critical to build a multi-year production model to accurately assess the MSC manufacturing requirements. This approach helps us make informed decisions about which manufacturing platform to select. In the example today, we have highlighted several platforms that are capable of meeting our production targets, but as we scale, our initial decisions may hinder our ability to move quickly to the next manufacturing milestone.”

Incidentally, manufacturing for extracellular vesicle (EV) therapies requires unique considerations beyond traditional cell therapies. Because EV production requires several-fold more cells to achieve target dose yields, one ought to consider scalable bioreactor systems, which can provide up to ten times the EV output compared to flask‑based methods. This key point reinforces the recurring theme of webinar: start with the right cell banks and scalable manufacturing platforms to set programs up for faster clinical progress and enduring success.

Question & Answer

With these key points established, the presenters then turned to some audience questions:

1. Why is it important to optimize the fill volume of vials when creating a working cell bank?
2. Are RoosterBio’s GMP materials already FDA‑approved or supported by regulatory filings?
3. Can the modeling approach shown in the webinar be applied to other cell types beyond MSCs?
4. Which bioreactor systems or scales were used in the modeling examples?
5. Where are RoosterBio’s GMP cell and media products manufactured?
6. How can a developer get started with 3D or bioreactor‑based manufacturing if they are currently using 2D systems?
7. What considerations are involved in fill‑finish and cryopreservation at larger manufacturing scales?

To be fully fortified with the answers to these questions, perhaps it’s best to access and consume the webinar and hear for yourself? RoosterBio looks forward to hearing from you when you have additional questions about how we and others approach these complicated issues.

Until we meet again for another webinar, we also invite you to noodle around with some model calculations to intuitively navigate production scale and dose via a downloadable Excel calculator tool, [2] related to a recent blog duet. [3, 4] Do we take pride in how we have helped our other customers establish their scalable bioprocess? …You can bank on it! [5, 6, 7, 8, 9]

  References

1. Snyder, Jessica, Carson, Jon. The Story Behind the Media. RoosterBio Blog 2022; Available from: https://www.roosterbio.com/blog/the-story-behind-the-media/.
2. RoosterBio. “One Does Not Simply Walk Into” a Therapeutic Cell Manufacturing Process… But the Long Road Need Not Be Perilous! RoosterBio Blog 2021; Available from: https://www.roosterbio.com/blog/one-does-not-simply-walk-into-a-therapeutic-cell-manufacturing-process-but-the-long-road-need-not-be-perilous/.
3. Lembong, Josephine, Rowley, Jon. Building Effective Multi-Year Process Development Programs I: Estimating hMSC Lot Size Ranges for Clinical Manufacturing Through Commercial Demand. RoosterBio Blog 2021; Available from: https://www.roosterbio.com/blog/building-effective-multi-year-process-development-programs-i/.
4. Lembong, Josephine, Rowley, Jon. Building Effective Multi-Year Process Development Programs II: Evolution of Technology Platform Decisions Based on Lot Size. RoosterBio Blog 2021; Available from: https://www.roosterbio.com/blog/building-effective-multi-year-process-development-programs-ii-evolution-of-technology-platform-decisions-based-on-lot-size/.
5. Lim, Mayasari. Know Your Cost of Goods Recap and Managing Lot Size Estimation. RoosterBio Blog 2020; Available from: https://www.roosterbio.com/blog/know-your-cost-of-goods-recap-and-managing-lot-size-estimation/.
6. RoosterBio. RoosterBio Galvanizes Large & Small for Clinical Success. RoosterBio Blog 2025; Available from: https://www.roosterbio.com/blog/roosterbio-galvanizes-large-small-for-clinical-success/.
7. RoosterBio. Leaving Flatland Behind – A View from Experience in 3D Adherent Cell Culture for Advanced Therapies. RoosterBio Blog 2024; Available from: https://www.roosterbio.com/blog/leaving-flatland-behind-a-view-from-experience-in-3d-adherent-cell-culture-for-advanced-therapies/.
8. RoosterBio. RoosterPubs in 2024 & Beyond: Catalyzing Bio-Research to “Publish & Flex”. RoosterBio Blog 2025; Available from: https://www.roosterbio.com/blog/roosterpubs-in-2024-beyond-catalyzing-bio-research-to-publish-flex/.
9. RoosterBio. Winners of RoosterBio’s 2024 Development Award. RoosterBio Blog 2025; Available from: https://www.roosterbio.com/blog/winners-of-roosterbios-2024-development-award/.