Revisiting “Identify & Define Your Cell Therapy’s Biomanufacturing Approach for Critical Process Parameters (CPPs)” by Josephine Lembong, PhD [1]

One of the most recycled mantras in cell and gene therapy manufacturing goes like this: “The process is the product.” Doesn’t that almost sound wise, like a lesson from Master Yoda (“…The product, the process is…”)? For unlike a small molecule drug whose chemical identity can be fully verified in a vial, a living cell therapy can never be completely characterized. Your manufacturing process is not just how you make the product. In one meaningful sense, it is the product.

Today, however, this pithy phrase invites some necessary scrutiny. As Professor Krishnendu Roy and colleagues articulated in their 2022 vision for next-generation cell manufacturing, [2] the field’s ultimate objective is more ambitious: the product should ultimately be the product. What does that riddle mean? The answer is simple. It means that a product is defined by what it’s made of and what it does, not just by the recipe that produced it. Reaching this confidence in product identity requires a well-honed process foundation rigorous enough to guarantee product quality, even when you can’t directly measure everything that matters. [3] Critical Process (CPPs) are the cornerstone of that foundation.

Figure 1. Above, a process overview for MSC expansion in 2D planar culture, and partial list of unit operations during passage. These tasks are associated with various CPPs, which potentially affect CQAs.

A Remedy for Unknown Causes of Known Headaches

A CPP is any process parameter — input or output — whose variability has a measurable impact on a downstream Critical Quality Attribute (CQA) and must therefore be monitored or controlled to ensure the process reliably produces a quality product. In plain language, a CPP is a dial on your manufacturing process that, if turned too far in either direction, produces a worse product. Figure 1 (above) shows us an example expansion process with mesenchymal stem/stromal cells (MSCs). An MSC process development expert’s job is to find those dials, understand how sensitive they are, and set acceptable ranges for each one.

What makes CPPs especially important in cell therapy is that some quality attributes cannot be measured with any current assay. These could include potency mechanisms not yet fully understood, or biological activities that only reveal themselves in human patients beyond cell culture or mice studies. These “unknown CQAs” (a concept developed by Dr. Mark Witcher and discussed in a prior blog series) can only be controlled through process consistency.

A well-defined CPP strategy is therefore not just required for successful IND filings. [4] It is the primary instrument for managing product quality you cannot directly see. And under FDA’s 2026 flexible CMC framework for CGTs, it’s also the currency you spend to earn the regulatory confidence that the new flexibility requires. [5]  With apologies to the cast of Monty Python in The Wind in the Willows, it may well be the Secret of Survival in a Very Nasty World of Cell and Gene Therapy Bioprocessing.

How to Find Your CPPs: Three Steps

Figure 2. Above, identification of CPPs comes down to three steps. It is impossible to evaluate the effect of the hundreds of interconnected parameters that exist within a cell and gene manufacturing process; hence, prioritization is important.

No development program has the resources to study each and every variable in a manufacturing process. A typical cell therapy workflow contains hundreds of interdependent parameters. Prioritization is the key skill for discernment of the practical approach. It comes down to three steps (Figure 2, above):

1. Map the process. Unit operations such as thaw, inoculation, expansion, harvest, formulation and fill each need to be documented. Enumerate all the known parameters within each step. This gap analysis is the foundation for everything that follows.
2. Quantify effects. Run targeted development studies to measure how each parameter of interest affects your defined CQAs. Tools that help you prioritize what to study include published literature, prior experience in related systems, and Failure Mode & Effects Analysis (FMEA). Where resources allow, Design of Experiment (DOE) approaches — which vary multiple parameters simultaneously in a planned pattern — are more efficient than one-at-a-time testing and will surface interaction effects that single-variable studies miss.
3. Define ranges. For those parameters that demonstrably affect CQAs, establish a target value and an acceptable operating range based on your experimental data. Set your target at the midpoint of the range. This builds in buffer against the variability you can’t fully control. As your process understanding grows, you can widen the range until you observe a failure point, which increases manufacturing tolerance without changing the process itself.

Input Parameters, Output Parameters, and Why the Difference Matters

Not all CPPs are the same. Some are inputs you set deliberately such as centrifuge speed, media composition, seeding density. Others are outputs that emerge from the process like the effective concentration of a growth factor after formulation, or cell density at harvest.

The distinction matters because the control strategy is different for each type. For an output parameter you can’t directly control in real time, try to avoid adding expensive in-process measurement. The right response is usually to run a development study establishing that CQAs are maintained across the full expected range of variability, then document that range and manage it through supplier qualification and material controls. Every additional measurement step adds cost, time, and potential for operator error. Build your control strategy to be as lean as the science permits.

The CPPs That Matter Most for MSC Bioreactor Expansion

When the three-step framework is applied to a 3D MSC expansion process, a consistent set of high-priority CPPs emerges (Figure 3, below). A recent systematic review and weighted Pareto analysis of bioreactor-based MSC expansion studies identified glucose and lactate concentration, headspace aeration, dissolved oxygen (DO), and pH as the parameters appearing most frequently as CPPs across the literature. They then reflected the well-documented effects on MSC proliferation, viability, and phenotype. [6]

Figure 3. Above, CPPs can be identified and monitored across the unit operations of 3D bioreactor expansion culture of MSCs and other primary cells. These parameters collectively govern the Critical Quality Attributes (CQAs) that define — and in early development, approximate — a cell therapy’s Target Product Profile (TPP), the desired clinical outcome the manufacturing process is ultimately built to deliver.

Cell inputs: seeding density, passage number, and PDL at inoculation determine the biological starting point of every batch. Maintaining a consistent PDL window is essential for batch-to-batch comparability. The same PDL and seeding density logic applies to human dermal fibroblast (hDF) expansion.

Media and raw materials: growth factor concentration and lot-to-lot media variability often behave as output-type parameters in practice, since supplier variability introduces ranges you don’t fully control. Development studies bracketing the expected range are the appropriate tool.

Bioreactor environment: pH, DO, temperature, and agitation rate each have demonstrated effects on MSC phenotype, proliferative capacity, and immunomodulatory function. DO deserves particular attention since culturing under physiological hypoxia (~3–5% O₂) versus atmospheric oxygen (~20%) can measurably improve expansion efficiency. [7, 8] Bioreactor shear stress from agitation is also meaningful. MSCs are mechanosensitive, and excessive shear can compromise viability and modify potency in ways that may not be immediately visible in standard release assays. [9]

Output parameters to track throughout: harvest density and viability, post-harvest recovery, and final PDL calculation collectively tell you whether the process performed as expected and provide the data needed to build your CPP<–>CQA design space over time.

One additional point is specific to MSC programs. That is, tissue source is a process-defining variable in its own right. Bone marrow, umbilical cord, and adipose-derived MSCs can differ in proliferative capacity, immunomodulatory potency, and differentiation potential. [10] Further, there are significant differences that are donor-dependent, irrespective of tissue origin. A GMP manufacturing template might accommodate multiple sources at equivalent yield, but the CQA and potency profiles of the resulting products are likely to differ. Thus, CPPs should be developed and validated around a single, specified tissue source.

The Bigger Picture

CPP control is the best available “console” for protecting quality that you cannot directly measure. Tightly controlled, reproducible processes are how the bioprocessing engineers manage unknown CQAs in the short-term outlook for the industry. Further, they are also the foundation from which better measurement is now emerging. As PAT tools, real-time in-process analytics, digital integration, and AI-assisted process modeling continue to mature, the structured, high-quality data generated by well-controlled CPP systems will become the training set that makes predictive manufacturing possible. This is beginning to close the loop between what is measured today and what can be guaranteed tomorrow.

Not incidentally, well-controlled CPPs are also the prerequisite for AI to work; predictive models are only as good as the process data they’re trained on. We are approaching an era where the product fully defines itself through analytics alone. But the CPP framework built rigorously today is what makes that future reachable.

At RoosterBio, our process development services are built to help you identify and control the “known unknowns” governing your CPPs and CQAs — from early characterization through GMP-ready manufacturing. Contact us to discuss your mesenchymal stem/stromal cell (M S C), human primary fibroblast, or EV/exosome program today.

 

References

1. Lembong, Josephine. Identify & Define Your Cell Therapy’s Biomanufacturing Approach for Critical Process Parameters (CPPs). 2021; Available from: https://www.roosterbio.com/blog/identify-define-your-cell-therapys-biomanufacturing-approach-for-critical-process-parameters-cpps/.
2. Chang, Sean, Bruce Greenwald, and Krish Roy, The digital revolution: technological innovations to enable automation in cell therapy manufacturing. Cell Gene Therapy Insights, 2022. 8: p. 355-369. 10.18609/cgti.2022.053
3. Campbell, Andrew, Thomas Brieva, Lior Raviv, Jon Rowley, Knut Niss, Harvey Brandwein, Steve Oh, and Ohad Karnieli. Concise Review: Process Development Considerations for Cell Therapy. Stem Cells Transl Med, 2015. 4(10): p. 1155-63. 10.5966/sctm.2014-0294
4. Administration, US Food and Drug. Potency assurance for cellular and gene therapy products. 2023; Available from: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/potency-assurance-cellular-and-gene-therapy-products
5. Administration, US Food and Drug. Flexible Requirements for Cell and Gene Therapies to Advance Innovation. 2026; Available from: https://www.fda.gov/vaccines-blood-biologics/cellular-gene-therapy-products/flexible-requirements-cell-and-gene-therapies-advance-innovation.
6. Herbst, L., B. Niessing, and R. H. Schmitt, Identification of critical process parameters and quality attributes for bioreactor-based expansion of human MSCs. Front Bioeng Biotechnol, 2025. 13: p. 1608194. 10.3389/fbioe.2025.1608194
7. Nikolits, I., et al., Towards Physiologic Culture Approaches to Improve Standard Cultivation of Mesenchymal Stem Cells. Cells, 2021. 10(4). 10.3390/cells10040886
8. Antebi, B., et al., Short-term physiological hypoxia potentiates the therapeutic function of mesenchymal stem cells. Stem Cell Res Ther, 2018. 9(1): p. 265. 10.1186/s13287-018-1007-x
9. Yu, H. and G. G. Hao, Bioreactors expansion of human mesenchymal stromal cell therapies: platforms, parameters, challenges and opportunities. J Genet Eng Biotechnol, 2026. 24(1): p. 100638. 10.1016/j.jgeb.2025.100638
10. Via, A. G., A. Frizziero, and F. Oliva, Biological properties of mesenchymal Stem Cells from different sources. Muscles Ligaments Tendons J, 2012. 2(3): p. 154-62.