Whether derived from autologous or allogeneic sources, advanced therapies deploy living organisms to target the root cause of disease rather than manage its symptoms. While manufacturing workflows may be well-defined on paper, the vein-to-vein logistics that underpin their clinical delivery are considerably more complex. Achieving a sufficient cell number is a necessary condition, but it is far from being the sole determinant of success.
A chimeric antigen receptor (CAR)-T therapy preparation begins with a median of 98 × 10⁸ nucleated cells, with the final dose containing 1-5 × 10⁶ cells per kilogram of patient’s body weight. © Green Elephant Biotech.
The term ‘advanced therapy medicinal products’ (ATMPs) was coined in 2003, as the prospects of regenerative medicine, repairing and treating with living tissue, could no longer be overlooked. Cell and gene therapies (CGTs) fell under this definition, building an extensive theoretical foundation over the following decades and differentiating between autologous and allogeneic sources of living material. While personalized treatments have generated considerable clinical success, allogeneic approaches have been developed in parallel for off-the-shelf application (1).
Although harmonized standards are still being established by regulatory agencies to support the clinical translation of ATMPs, the foundational steps and workflows of therapy development are well-defined. From cell isolation and seed train culture initiation through to expansion, safe and effective CGTs, including those produced with viral vectors, are no longer conceptual. However, out of the 40+ living therapies approved by the Food and Drug Administration (FDA), a notable proportion received authorization as far back as the early-to-mid 2000s (2).
Available data indicates that even for chimeric antigen receptor (CAR)-T cells — an immunotherapy with well-documented clinical success since 2017 and established suspension-based manufacturing, approximately 7-8 out of every 10 eligible patients do not ultimately receive treatment. While patient hesitancy is a contributing factor, operational and financial burdens remain significant, compounded by the scale of cell numbers involved in therapy development. This environment underscores the importance of engaging all stakeholders around these challenges (2).
This entry brings together published data and real-world figures to examine just how many cells are required to produce a CGT, including whether current infrastructure is well-equipped to handle the objective. Importantly, the numbers presented represent order-of-magnitude ranges, as actual figures vary considerably depending on the manufacturing platform, therapy indication, and dosing. Improved protocols and technological advancements hold the potential to refine these estimates further, supporting the bench-to-bedside translation of CGTs and broadening patient access.
From leukapheresis to therapeutic dose: Just how many cells does a CGT require?
As noted above, autologous therapies require an additional step – cell isolation directly from the patient. In the context of CAR-T cells, for instance, a standard leukapheresis procedure, the apheresis-based removal of white cells from the bloodstream, yields a median of 98 × 10⁸ nucleated cells with a viability of 99.6-100% in a volume of approximately 237 mL. Ideally, sample collection is a single procedure, but depending on the patient’s condition and available cell count, multiple rounds may be necessary. Clinical evidence also suggests that smaller blood volumes of 50-100 mL could be sufficient, provided CAR-T cells achieve at least 100-fold expansion within 10-20 days, though this remains a subject of ongoing debate (3).
Following cell collection, the optimal number for culture initiation varies by cell type and available resources, with one report recommending 50-100 × 10⁶ cells as a starting range (4). At this stage, a distinction emerges between transfection and transduction. Transfection employs non-viral delivery tools such as lipid nanoparticles to introduce genetic material into cells. 79-90% confluency for adherent cells or 5 × 10⁵-2 × 10⁶ cells/mL for suspension cultures are generally regarded as indicators of successful transfection. Transduction, which relies on a viral vector, follows a different mechanism, with efficiencies typically ranging between 30-70% — a spectrum influenced by familiar variables including cell type, health status, and viability (5, 6).
To reach a therapeutic cell density using genetically engineered cells, the expansion process is supported by growth factor supplementation tailored to the cell’s identity and therapeutic purpose. For a CAR-T-based ATMP, the expansion phase requires approximately 100 million cells, and for 99% of patients, the product is ready within 9-10 days. Studies have indicated the possibility of compressing CAR-T cell therapy manufacturing to as little as a few days or even 24 hours, although such acceleration is not standard practice as of today. The variation in production timelines is thought to reflect differences in the initial cell count available, transduction efficiency, and expansion rate (7, 8).
Not all CGT types, however, are satisfied with millions of cells. Autologous tumor-infiltrating lymphocytes are expanded into the billions, comparable to allogeneic therapy manufacturing, while iPSC-derived cell doses may be even higher. For mesenchymal stromal cells (MSCs) — one of the most widely used adherent cell types in advanced therapies, a typical therapeutic dose ranges from 20 to 200 × 10⁶ cells per treatment. The cell counts known to produce a clinically meaningful response are informed by clinical trial data, and are closely monitored at key cell therapy manufacturing steps through cell-based assays and imaging where applicable (9, 7, 3).
Real-life adherent showcases that illuminate the difference between millions and billions of cells
To illustrate the range of cell numbers applied across CGTs, two approved ATMPs with well-defined dosing regimens are examined below, with brief reference to their clinical indications and outcomes. An important contextual consideration is the distinction between autologous and allogeneic treatment, as the latter can be mass-produced at scale, reducing cost per dose and supporting standardized quality. With the autologous model defined by its personalized nature, lower cell volumes reflect not only biological requirements, as described below, but also the constraints of demand and individual applicability (8).
MACI® (autologous cultured chondrocytes on porcine collagen membrane) represents a different therapeutic modality compared to infused CAR-T products. Indicated for cartilage defects of the knee in skeletally mature patients, anchorage-dependent chondrocytes are isolated at densities ranging from 10,400 to 1.07 × 10⁶ cells per cm². Following ex vivo expansion, the implantation matrix is seeded with chondrocytes at a density of 500,000-1,000,000 cells per cm² onto a type I/III porcine-derived collagen membrane. The randomized parallel-group clinical trial supporting the ATMP’s approval enrolled 144 patients and recorded an adverse reaction rate of only 5%, providing a compelling evidence base for clinical success. Nevertheless, marketing authorization for MACI® in Europe was subsequently suspended due to the absence of an authorized manufacturing site and batch release (10, 11, 12).
By contrast, many allogeneic therapies already carry the intrinsic capacity to address the target pathology and frequently do not require genetic re-engineering as part of the cell therapy manufacturing process. In the case of Ryoncil® (remestemcel-L-rknd), an MSC-based CGT, donor-derived MSCs are employed to attenuate inflammation through their native immunomodulatory mechanisms. In pediatric recipients with steroid-refractory acute graft-versus-host disease, the therapy produced a clinically meaningful response in 70% of cases, enabling patients to discontinue immunosuppression within less than a year of treatment. Ryoncil® is administered twice weekly over the course of approximately one month at a dose of 2 × 10⁶ cells per kilogram of body weight — a regimen designed to achieve a durable response, supported by the well-established safety profile of MSCs (13, 14).
The cell numbers required across CGT development therefore reflect a spectrum of interdependent factors, including the final product’s affinity, clinical indication, and available manufacturing infrastructure. Manufacturing of anchorage-dependent cells is frequently associated with technically demanding and labor-intensive processes. MSCs, for instance, are typically seeded at a standard density of 4,000 cells per cm², with harvest targets of 30,000–50,000 cells per cm² recommended to prevent nutrient depletion and metabolic waste accumulation. Prescribing information for both MACI® and Ryoncil® indicates that a single dose requires anywhere from 16,000,000 to 60,000,000 cells — a threshold that must be met repeatedly in most allogeneic therapy regimens (Table 1). Reaching these harvest densities demands a minimum of eight T-175 flasks per run, each introducing the well-documented risks of uneven cell distribution and pathogenic contamination. Whether the target dose is measured in millions or billions, the need for robust, standardized manufacturing workflows will remain a defining challenge across ATMPs until meaningfully addressed (15).
| Therapy name | Company | Approval year | Dose | No. of cells in 1 dose | Dose frequency | App. surface area per dose (in cm²) | Vessel count (T-175) per dose | Vessel count (CellScrew®) per dose |
| MACI® (autologous cultured chondrocytes on porcine collagen membrane) | Vericel Corp. | 2016 | 0.5-1 × 10⁶ chondrocytes/cm² | For an average implantation matrix of ~16 cm²: 16,000,000 cells | 1 | 320 | 4 | 1 CellScrew® mini |
| Ryoncil® (remestemcel-L-rknd) | Mesoblast Ltd. | 2024 | 2 × 10⁶ MSCs/kg of body weight | For the pediatric patient’s body weight of ~30 kg: 60,000,000 | 8+ | 1.200 | 6 | 2 CellScrew® mini or 1 CellScrew® 6K |
| Prochymal (remestemcel-L) | Mesoblast Ltd. | 2012 | 2 × 10⁶ hMSCs/kg of body weight | For the pediatric patient’s body weight of ~30 kg: 60,000,000 | 4 | 1.200 | 6 | 2 CellScrew® mini or 1 CellScrew® 6K |
| Alofisel® (darvadstrocel) | TiGenix NV/Takeda | 2018 | 120 × 10⁶ MSCs | A fixed injection of 120,000,000 cells | 1 | 2.400 | 13 | 3 CellScrew® mini or 1 CellScrew® 6K |
| Kymriah® (tisagenlecleucel-T) | Novartis Pharmaceutical Corp. | 2017 | 0.1-2.5 × 10⁸ of CAR-T cells | For the patient’s body weight of ~50+ kg: A fixed injection of 250,000,000 cells | 1 | 5.000 | 29 | 1 CellScrew® 6K |
| Ebvallo® (tabelecleucel) | Atara Biotherapeutics | 2022 | 2 × 10⁶ T cells/kg of body weight | For the patient’s weight of ~80 kg: 160,000,000 | 3 | 3.200 | 18 | 1 CellScrew® 6K |
The complexity of transitioning from seed train to bioreactor also shapes patient access
The objective behind manufacturing any drug at a given cell density is anchored in three principles: consistency, cost-efficiency, and scalability. Regulatory and manufacturing frameworks, however, are still evolving to accommodate the specific demands of living cell products. Seed train culture expansion, for instance, is a well-mapped procedure in which approximately 5-20 × 10⁶ cells from a working cell bank cryovial undergo a 30,000-60,000-fold expansion to eventually seed a 2,000 L bioreactor. Beyond the contamination risks inherent to 5-8 sequential cell passages, declining viability and reduced growth rates represent challenges that are largely specific to research-stage and CGT development workflows (18).
When cultivating cells at high density, regulatory agencies recognize that certain parameters must retain a degree of flexibility. For example, any two CAR-T cell therapy products characterized as “>90% CAR+ T cells” by the same manufacturer are likely to differ meaningfully in composition — each containing different proportions of cytotoxic, memory, and helper T cells. While this variability rarely compromises purity, the consistency and potency of the therapy become increasingly uncertain as cell subsets remain uncharacterized or undeclared within the identity panel (15).
Cell exhaustion driven by lactate accumulation or other metabolite buildup represents a further primary bottleneck in high-density cell cultivation. Lactate concentrations exceeding 7.0 g/L have been shown to contribute to cell death by reducing culture pH and disrupting normal cellular function. This challenge, relatively specific to CGT workflows, limits the utility of static cell cultures and legacy fed-batch systems when expanding large cell volumes. Its significance is compounded by the fact that expansion is a rate-limiting step in both autologous and allogeneic therapy production (20).
The real-world consequences of these manufacturing constraints fall directly on patients. Scientific innovation outpacing infrastructure is not unprecedented, but predictable access to therapies of consistent quality remains the exception rather than the rule. Since the completion of the Human Genome Project in 2003, investment in the CGT space has reached 15.2 billion USD in 2024 alone, meanwhile only a small fraction of candidates achieve commercial authorization. The causes behind this gap are numerous, but manufacturing bottlenecks inflate the cost of clinical trial failure, ultimately driving up treatment prices and reducing access for the patients who need these therapies most (21).
The figures behind cell and gene therapy development tell a precise story. From the median 98 × 10⁸ nucleated cells obtained at leukapheresis to the 30,000-60,000-fold expansion required to seed a commercial bioreactor, every order of magnitude carries weight. ATMPs like MACI® and Ryoncil® illustrate that the right cell count does not have to be the largest one, but one that balances therapeutic efficiency with patient safety and process reproducibility. This principle extends onto emerging therapeutic interventions, such as CAR-macrophages (CAR-M), where the issues of short cell life and fragile handling are already recognized to hinder the traditional scale-up strategies.
In the field where manufacturing bottlenecks interfere with the returns from 15.2 billion USD investments and the chances of 7-8 out of 10 eligible candidates of receiving treatment, the gap emerges as a concern to be addressed urgently. Today, standardized, economically viable, and scalable cell cultivation systems are prerequisite for translating biological promise into therapies patients can reliably access.
Path to patient access supported by Archimedes® One
The Archimedes® One dynamic adherent bioreactor addresses the core manufacturing objectives of scalability, consistency, and cost-efficiency through automation built around the CellScrew® culture system. Supporting cell densities of up to 7.7 × 10⁵ cells per cm² for adherent cell lines such as HEK-293, Vero B4, and MSCs, the system facilitates robust cell expansion while maintaining viability consistently above 92%. The efficiency gains extend directly to resource consumption. Mechanical harvesting in CellScrew® workflows requires 67% less detachment reagent, reducing both reagent-related costs and the enzymatic exposure, combined with up to 80% less material utilization during the vessel’s manufacturing.
If you are looking to advance your cell therapy manufacturing workflow — whether through cultivation data or platform consultation, the Green Elephant Biotech page has a range of resources for personalized support.
References:
- Perin, F, Ouyang, L, Lim, KS, Motta, A, Maniglio, D, Moroni, L, Mota, C, Bioprinted Constructs in the Regulatory Landscape: Current State and Future Perspectives (2025), Adv Mater 38: 1-22. doi: 10.1002/adma.202504037.
- Is Community the Next Frontier for Cell and Gene Therapy? (2025). Elias C Pittos, Robert E Richards, Rocky Billups, Pharmaceutical Executive. Available https://www.pharmexec.com/view/community-next-frontier-cell-gene-therapy (Accessed 22 July 2026).
- Han, L, Zhou, J, Li, L, Zhou, K, Zhao, L, Zhu, X, Yin, Q, Li, Y, You, H, Zhang, J, Song, Y, Gao, Q, Culturing adequate CAR-T cells from less peripheral blood to treat B-cell malignancies (2021), Cancer Biol Med 18: 1066-79. doi: 10.20892/j.issn.2095-3941.2021.0040.
- Dias, J, Garcia, J, Agliardi, G, Roddie C, CAR-T cell manufacturing landscape — Lessons from the past decade and considerations for early clinical development (2024), Mol Ther Methods Clin Dev 32: 1-17. doi: 10.1016/j.omtm.2024.101250.
- Dan, L, Kang-Zheng, L, Optimizing viral transduction in immune cell therapy manufacturing: key process design considerations (2025), J Transl Med 23: 1-17. doi: 10.1186/s12967-025-06524-0.
- Factors Influencing Transfection Efficiency. (2026). ThermoFisher Scientific. Available https://www.thermofisher.com/de/de/home/references/gibco-cell-culture-basics/transfection-basics/factors-influencing-transfection-efficiency (Accessed 16 July 2026).
- Cell Expansion: The Rate Limiting Step for Cell Therapies. (2022). Angelo DePalma, Biocompare. Available https://www.biocompare.com/Editorial-Articles/585310-Cell-Expansion-The-Rate-Limiting-Step-for-Cell-Therapies/ (Accessed 16 July 2026).
- Pessach, I, Nagler, A, Leukapheresis for CAR-T cell production and therapy (2023), Transfus Apher Sci 62: 1-13. doi: 10.1016/j.transci.2023.103828.
- Stem Cell Dosage in Regenerative Medicine: How Much Is Enough? (2025). Vibrant Health Clinics. Available https://www.vibranthcs.com/blog/stem-cell-dosage-in-regenerative-medicine (Accessed 07 August 2026).
- MACI®. (2024). John P Cunha, RxList. Available https://www.rxlist.com/maci-drug.htm (Accessed 27 July 2026).
- Jong, WC, Penn, TW, Single-Surgery Co-Implantation of Autologous Primary Articular Chondrocytes with Bone Marrow Cells in the Treatment of Articular Cartilage Lesions: Observations from the Operating Room on Tissue Input and Cell Output (2022), Cartilage 13: 32-42. doi: 10.1177/19476035221132259.
- Maci®, matrix-applied characterised autologous cultured chondrocytes. (2018). European Medicinal Agency. Available https://www.ema.europa.eu/en/medicines/human/EPAR/maci (Accessed 27 July 2026).
- Etra, A, Ferrara, JLM, Levine, JE, Ryoncil (remestemcel-L-rknd): the first approved cellular therapy for steroid-refractory acute GVHD (2025), Blood 146: 1897-1901. doi: 10.1182/blood.2025028553.
- RYONCIL® (remestemcel-L-rknd) suspension for intravenous infusion. (2024). The Food and Drug Administration. Available https://www.fda.gov/media/184603/download (Accessed 27 July 2026).
- Lowdell, MW, Considerations for manufacturing of cell and gene medicines for clinical development (2025), Cytother 27: 874-83. doi: 10.1016/j.jcyt.2024.11.015.
- Han, X, Liao, R, Li, X, Zhang, C, Huo, S, Qui, L, Xong, Y, He, T, Xiao, G, Zhang, T, Mesenchymal stem cells in treating human diseases: molecular mechanisms and clinical studies (2025), Sig Tranduct Target Ther 10: 1-41. doi: 10.1038/s41392-025-02313-9.
- Kymriah® (tisagenlecleucel-T). (2025). The Food and Drug Administration. Available https://www.fda.gov/media/107296/download (Accessed 23 July 2026).
- Ebvallo® (tabelecleucel). (2023). European Medicinal Agency. Available https://www.ema.europa.eu/en/medicines/human/EPAR/ebvallo (Accessed 23 July 2026).
- How to Develop a Seed Train for Mammalian Cell Culture. (2026). BioPrcoess Tools. Available https://bioprocesstools.com/blog/seed-train-development/ (Accessed 21 July 2026).
- Demonstrating Scalable T-Cell Expansion in Stirred-Tank Bioreactors. (2018). Alex Klarer, David Smith, Ryan Cassidy, Thomas Heathman, Qasim Rafiq, BioProcess International. Available https://www.bioprocessintl.com/cell-therapies/demonstrating-scalable-t-cell-expansion-in-stirred-tank-bioreactors (Accessed 21 July 2026).
- Understanding the ATMP Revolution in Life Sciences. (2022). Nicholas Basta, Leading Minds Network. Available https://www.leading-minds-network.com/en/article/understanding-atmp-revolution-in-life-sciences (Accessed 21 July 2026).
