医院内 CAR-T 生产网络如何搭建，作者分析设施成本与英欧美监管路径

By Boro Dropulić, Ph.D., MBA, Caring Cross

Implementing point-of-care (POC) CAR T cell therapy manufacturing hinges on a distributed network of sites maintaining manufacturing and quality systems under a shared, network-wide framework.

Part 1 of this series showed that POC manufacturing, defined as production within or adjacent to the treating hospital under a licensed quality system, can deliver products with efficacy and safety comparable to centralized sites at a fraction of the cost.

GMP Facilities: Scale, Utilization, And Process Design

A study modeling the CAR T supply chain showed that centralized manufacturing gains economies of scale, whereas smaller decentralized facilities, although each producing far fewer products, used their resources more efficiently, bringing their cost per treatment close to that of centralized manufacturing.¹ In practice, demand at any single site is rarely constant, which favors smaller, flexible facilities. Efficiency is further increased when hospital networks share process development, quality systems, training, regulatory filings, and vector supply, so those fixed costs are borne once.

This is part 2 of a series on improving access to CAR T cell therapy through distributed manufacturing. If you haven't read it yet, see part 1 here

Small GMP facilities give manufacturers a cost-effective entry point into CAR T cell manufacturing, because unit cost depends not only on materials and labor but on how efficiently a facility is used.² Modular workflows make the most of a small footprint: compact, functionally closed cell-processing instruments (e.g., Sepax C-Pro, Rotea) for cell isolation and washing^3,4 and gas-permeable culture vessels (such as G-Rex or gas-permeable culture bags) for expansion in standard incubators^5,6 allow several products at different stages to be manipulated in the same room at different times. With such a process, it is estimated that a small GMP facility can start producing one patient-specific therapy per day using a seven-day process, or more than 250 products per year, from a single small room on a single labor shift. Integrated automated devices, by contrast, have high capital and material costs and require dedicated single-source consumables. Each device also occupies its own floor space and is dedicated to a single product for the full seven- to 14-day process, so throughput per unit of cleanroom space is substantially lower, and adding capacity requires buying more instruments and adding more space.⁷ Modular workflows therefore offer several advantages over integrated devices: they need less floor space and use GMP space more efficiently, which lowers the cost per product;⁸ they carry lower capital and maintenance costs;⁷ they can draw on consumables from multiple suppliers, avoiding the supply risk, vendor lock-in, and price increases that can come with sole-source materials;⁹ and each step can be upgraded or replaced independently as processing methods improve, without depending on a single device manufacturer.¹⁰ Radio-frequency identification (RFID) tracking of each patient's product can make fuller use of GMP space by allowing several products to be manufactured safely in the same small cleanroom, as is already done in in vitro fertilization (IVF) laboratories, where RFID witnessing detected all simulated sample mix-ups compared with 88.7% for printed labels.¹¹

Combining small GMP facilities with modular workflows lowers the capital needed to establish CAR T cell production capability,⁷ so manufacturers can start with a modest, financially viable operation and add capacity as demand grows, rather than risk building and maintaining a large, centralized facility based on integrated devices ahead of demand.¹²

Innovation Incentives

Patents on currently marketed medicines representing more than $236 billion in sales are expected to expire between 2025 and 2030,¹³ and cell and gene therapies are among the modalities on which the next generation of therapeutic innovation depends. Industry groups have expressed concern that the EU hospital exemption, which allows advanced therapies to be prepared on a non-routine basis and used in a hospital in the same member state for an individual patient without marketing authorization, could weaken investment incentives.¹⁴ Yet innovation in cellular gene therapy has come largely from universities and academic medical centers, where most approved CAR T products originated, including tisagenlecleucel (Kymriah) at the University of Pennsylvania¹⁵ and axicabtagene ciloleucel (Yescarta) at the U.S. National Cancer Institute.¹⁶ Because a new CAR T product is created mainly by redesigning its genetic construct, each round of innovation costs far less than discovering and developing a new drug molecule. Such innovation will therefore continue regardless of the manufacturing model, and a POC model may even accelerate it by increasing the number of investigator-initiated trials, which cost substantially less than company-sponsored trials.¹⁷ Companies cannot always provide therapies that patients urgently need when the economics do not create sufficient shareholder value. GD2-directed CAR T cells are a clear example: academic trials have shown encouraging results in children with high-risk neuroblastoma¹⁸ and H3K27M-mutated diffuse midline glioma,^19,20 an almost uniformly fatal brain cancer, yet no anti-GD2 CAR T product has received marketing authorization. A POC manufacturing network can fill this gap.

Regulation, Hospital Safety Nets, And National Capacity

The U.K. Human Medicines (Modular Manufacture and Point of Care) Regulations 2025 require a named control site and a master file for each product.²¹ In the U.S., a proposed FDA rule published in July 2026 would allow a distributed manufacturing hub and its manufacturing units to register as a single establishment.²² In the EU, the reformed pharmaceutical legislation, adopted by the council in September 2026 and pending final approval by the European Parliament, would introduce provisions on decentralized manufacturing.²³

Open aseptic processing requires cleanrooms with Grade A operations in a Grade B background and Grade C areas for supporting steps. Closed systems, however, may be operated in a Grade D environment under EU GMP guidelines for advanced therapies,²⁴ which also apply in the U.K.,²⁵ substantially reducing construction and operating costs. The Pharmaceutical Inspection Co-operation Scheme (PIC/S) GMP guide, used by regulators in more than 50 countries, similarly bases area classification on whether processing is open or closed,²⁶ and both EU GMP Annex 1 and U.S. FDA guidance recognize the lower contamination risk of closed-system processing, including isolators.^27,28 Closed-system culture platforms, such as gas-permeable culture vessels, support this approach. Alternatives to conventional cleanrooms include prefabricated biomodular cleanroom units, such as those produced by Germfree,²⁹ and isolator workstations, which can be operated in a Grade D background.³⁰

Within these frameworks, hospitals capable of POC manufacture can serve as a safety net for patients when commercial products are unavailable, delayed by slot allocation or payer authorization, or offered too far from the patient.^31,32,33 Products can progress through a staged pathway: investigator-initiated trials, early access under national mechanisms such as expanded access and hospital exemption, then pivotal studies supporting a marketing application, and finally full authorization and reimbursement. The ARI-0001 program illustrates this pathway: it was developed at Hospital Clínic Barcelona, approved under the Spanish hospital exemption, and is now reimbursed nationally.^34,35 These frameworks could also become open to commercial developers, who could distribute their own products through networks of POC sites.^21,22 Hospital manufacturers carry the same pharmacovigilance obligations, including the lifelong monitoring for secondary malignancies now specified for all approved CAR T products,³⁶ and can meet them through the registries already used for transplantation and cellular therapy. POC manufacturing does not by itself remove clinical constraints: accredited cellular therapy units, intensive-care capacity, and specialist clinical staff remain prerequisites, so it is best placed initially at centers that already deliver hematopoietic stem cell transplantation. The FDA's 2025 removal of REMS requirements for autologous CAR T products and new FACT standards for community centers are already widening the set of sites that can administer these therapies.^37,38

Local Manufacturing Builds National Resilience

Solving the global access problem for cellular gene therapies will require broad transfer of this technology so that countries can produce these therapies locally rather than depend on imports from a small number of multinational manufacturers. The United States has withdrawn from the World Health Organization,³⁹ and multinational companies have been unable to supply advanced therapies to LMICs at affordable prices. At a side event during the 2026 World Health Assembly in Geneva, Indonesian Minister of Health Budi Gunadi Sadikin said: "National resilience of the healthcare supply chain infrastructure through local manufacturing is far more critical in this new era than cost efficiency alone."⁴⁰

The economic benefits of local manufacture extend well beyond the price of a single product. Local labor and simpler logistics substantially reduce the per-patient cost of therapy, while investment in CAR T manufacturing builds facilities and expertise that can be applied to future health needs. Local production also strengthens supply chain security, reducing reliance on foreign suppliers and volatile global markets. A domestic cell therapy sector creates high-value positions for scientists, engineers, and clinicians, retains local talent, and can attract regional medical tourism. Local capacity also allows each country to develop therapies for the diseases most relevant to its own population.

Hospital-Based Production Fills The Gaps In Commercial Supply

POC manufacturing is an important safety net that complements existing commercial models. Prospective studies show that hospital-manufactured CAR T products achieve efficacy and safety comparable to commercial products, and short on-site processes can reduce vein-to-vein time by several weeks, which is critical for patients with advanced disease. Affordable vector, closed-system processing, and efficient use of small GMP facilities reduce costs several-fold, and further still where skilled labor is less expensive.

Hospital networks with shared intelligent quality systems, closed manufacturing systems, and digital real-time tracking can address the remaining challenges of quality control, utilization, and comparability, while a staged regulatory pathway allows hospital-manufactured products to reach full approval. Where commercial supply is delayed, unaffordable, or absent, POC manufacturing enables hospitals to treat patients who would otherwise go untreated. Commercial developers will continue to bring branded, patent-protected therapies to market; hospital-based manufacturing can ensure access to nonbranded therapies wherever gaps remain. Together, they can substantially improve access to these often lifesaving therapies for patients around the world.

Editor's note: The author discloses that he is executive director of Caring Cross and chief executive officer of Vector BioMed. Caring Cross has collaborations with Germfree, Wilson Wolf, and Cytiva, whose products are mentioned in this article.

References:

1. Lam C, Meinert E, Yang A, Cui Z. Comparison between centralised and decentralised supply chain of autologous CAR T-cell therapies. Cytotherapy 2021;23(5):433–451. doi:10.1016/j.jcyt.2020.08.007
2. Ran T, Eichmüller SB, Schmidt P, Schlander M. Cost of decentralized CAR T-cell production in an academic nonprofit setting. Int J Cancer 2020;147(12):3438–3445. doi:10.1002/ijc.33156
3. Cytiva. Sepax C-Pro cell processing system. Accessed October 2026. https://info.cytivalifesciences.com/automated-cell-processing.html
4. Thermo Fisher Scientific. Gibco CTS Rotea Counterflow Centrifugation System. Accessed October 2026. https://www.thermofisher.com/rotea
5. ScaleReady. ScaleReady Platform: G-Rex CGT manufacturing; G-Rex for CAR-T cell therapy. Accessed October 2026. https://www.scaleready.com/scaleready-platform/
6. van Schalkwyk MCI, van der Stegen SJC, Bosshard-Carter L, et al. Development and validation of a good manufacturing process for IL-4-driven expansion of chimeric cytokine receptor-expressing CAR T-cells. Cells 2021;10(7):1797. doi:10.3390/cells10071797
7. Shah M, Krull A, O'Donnell L, de Lima MJ, Bezerra E. Promises and challenges of a decentralized CAR T-cell manufacturing model. Front Transplant 2023;2:1238535. doi:10.3389/frtra.2023.1238535
8. Odum J. Optimizing cost of goods for cell therapy manufacturing. Pharmaceutical Engineering (ISPE), November/December 2023. https://ispe.org/pharmaceutical-engineering/november-december-2023/optimizing-cost-goods-cell-therapy-manufacturing
9. DiGiusto D, Leistler B, Walls T. Overcoming raw material challenges in cell & gene therapy manufacturing. Cell Gene Ther Insights 2020;6(2):415–426. doi:10.18609/cgti.2020.050
10. Ji Y. Applying a closed, modular, semi-automated system to CAR T cell therapy manufacturing. Cell Gene Ther Insights 2021;7(8):961–972. doi:10.18609/cgti.2021.127
11. Jiang MX, Guo L, Li S, et al. IVF laboratory management through workflow-based RFID tag witnessing and real-time information entry. Reprod Biol Endocrinol 2024;22:96. doi:10.1186/s12958-024-01267-x
12. Harrison RP, Medcalf N, Rafiq QA. Cell therapy-processing economics: small-scale microfactories as a stepping stone toward large-scale macrofactories. Regen Med 2018;13(2):159–173. doi:10.2217/rme-2017-0103
13. BioStock. USD 236 billion at stake as patent collapse approaches. March 2025. https://biostock.se/en/2025/03/236-miljarder-usd-pa-spel-nar-patentstupet-narmar-sig/
14. ARM, EFPIA, EuropaBio, EUCOPE, ISCT. Joint position paper on the hospital exemption scheme for ATMPs. July 2023. https://www.europabio.org/reform-of-the-eu-general-pharmaceutical-legislation-joint-position-paper-on-the-hospital-exemption-scheme-for-atmps/
15. Penn Medicine. Penn Medicine-developed CAR T cell therapy wins third FDA approval. 31 May 2022. https://www.newswise.com/articles/penn-medicine-developed-car-t-cell-therapy-wins-third-fda-approval
16. National Cancer Institute. FDA approval of axicabtagene ciloleucel (Yescarta) for lymphoma. Cancer Currents blog, 25 October 2017. https://www.cancer.gov/news-events/cancer-currents-blog/2017/yescarta-fda-lymphoma
17. Paul M, Harbarth S, Huttner A, et al. Investigator-initiated randomized controlled trials in infectious diseases: better value for money for registration trials of new antimicrobials. Clin Infect Dis 2021;72(7):1259–1264. doi:10.1093/cid/ciaa930
18. Del Bufalo F, De Angelis B, Caruana I, et al. GD2-CART01 for relapsed or refractory high-risk neuroblastoma. N Engl J Med 2023;388(14):1284–1295. doi:10.1056/NEJMoa2210859
19. Majzner RG, Ramakrishna S, Yeom KW, et al. GD2-CAR T cell therapy for H3K27M-mutated diffuse midline gliomas. Nature 2022;603(7903):934–941. doi:10.1038/s41586-022-04489-4
20. Monje M, Mahdi J, Majzner R, et al. Intravenous and intracranial GD2-CAR T cells for H3K27M+ diffuse midline gliomas. Nature 2025;637(8046):708–715. doi:10.1038/s41586-024-08171-9
21. Medicines and Healthcare products Regulatory Agency. Human Medicines (Modular Manufacture and Point of Care) Regulations 2025: overview. 2025. https://www.gov.uk/government/publications/human-medicines-modular-manufacture-and-point-of-care-regulations-2025-overview
22. US Food and Drug Administration. Drug establishment registration and drug listing requirements for establishments engaged in distributed manufacturing and certain foreign establishments (proposed rule). Fed Regist 2026;91:42888. Summary: https://www.arnoldporter.com/en/perspectives/advisories/2026/07/fda-proposes-new-registration-and-listing-pathway-for-distributed-manufacturing
23. European Parliament and Council provisional agreement on the EU pharmaceutical package, 11 December 2025; Council first-reading position adopted 28 September 2026, European Parliament vote pending. Summaries: https://www.gmp-compliance.org/gmp-news/eu-pharma-package-council-adopts-new-rules; https://www.potterclarkson.com/news/the-european-parliament-and-council-reach-provisional-agreement-on-the-eu-pharma-package
24. European Commission. EudraLex Volume 4, Part IV: Guidelines on Good Manufacturing Practice specific to Advanced Therapy Medicinal Products, C(2017) 7694 final. 22 November 2017; paragraph 9.41. https://health.ec.europa.eu/document/download/ad33d9dd-03f0-4bef-af53-21308ce2187d_en
25. Medicines and Healthcare products Regulatory Agency. Freedom of information request FOI 22/1125: GMP guidelines applicable to advanced therapy medicinal products. 2022. https://www.gov.uk/government/publications/freedom-of-information-responses-from-the-mhra-week-commencing-28-november-2022/freedom-of-information-request-foi-221125
26. Pharmaceutical Inspection Co-operation Scheme (PIC/S). Guide to Good Manufacturing Practice for Medicinal Products, Annex 2A: Manufacture of advanced therapy medicinal products for human use. PE 009-15, 2021. Summary: https://www.pharmout.net/annex-2a-manufacture-of-atmps-pics/
27. European Commission. EudraLex Volume 4, Annex 1: Manufacture of Sterile Medicinal Products. C(2022) 5938 final. 22 August 2022; paragraph 8.127. https://health.ec.europa.eu/system/files/2022-08/20220825_gmp-an1_en_0.pdf
28. US Food and Drug Administration. Guidance for industry: Sterile drug products produced by aseptic processing — current good manufacturing practice. September 2004. https://www.fda.gov/media/71026/download
29. Germfree launches mobile cleanroom technology for CGTs. Cleanroom Technology, 8 February 2022. https://www.cleanroomtechnology.com/news/article_page/Germfree_launches_mobile_cleanroom
    _technology_for_CGTs/198657
30. Chiu YS, Wang KL, Wu KL. Isolator-based point-of-care manufacturing: a practical solution for GMP-compliant cell and extracellular vesicles therapy production. Front Bioeng Biotechnol 2025;13:1644318. doi:10.3389/fbioe.2025.1644318
31. Chung AP, Shafrin JT, Vadgama S, et al. Inequalities in CAR T-cell therapy access for US patients with R/R DLBCL: a SEER-Medicare analysis. Blood Adv 2025;9(18):4727–4735. doi:10.1182/bloodadvances.2024015634
32. Ahmed N, Wesson W, Mushtaq MU, et al. "Waitlist mortality" is high for myeloma patients with limited access to BCMA therapy. Front Oncol 2023;13:1206715. doi:10.3389/fonc.2023.1206715
33. Snyder S, Chung KC, Jun MP, et al. Access to chimeric antigen receptor T cell therapy for diffuse large B cell lymphoma. Adv Ther 2021;38(9):4659–4674. doi:10.1007/s12325-021-01838-z
34. Trias E, Juan M, Urbano-Ispizua A, Calvo G. The hospital exemption pathway for the approval of advanced therapy medicinal products: an underused opportunity? The case of the CAR-T ARI-0001. Bone Marrow Transplant 2022;57(2):156–159. doi:10.1038/s41409-021-01463-y
35. Ortiz-Maldonado V, Martinez-Cibrian N, Alserawan L, et al. Varnimcabtagene autoleucel for adults with relapsed or refractory B-cell precursor ALL (CART19-BE-02). Lancet Haematol 2026;13(3):e133–e143. doi:10.1016/S2352-3026(25)00328-X
36. US Food and Drug Administration. FDA requires boxed warning for T cell malignancies following treatment with BCMA-directed or CD19-directed autologous chimeric antigen receptor (CAR) T cell immunotherapies. April 2024. https://www.fda.gov/vaccines-blood-biologics/safety-availability-biologics/fda-requires-boxed-warning-t-cell-malignancies-following-treatment-bcma-directed-or-cd19-directed
37. Heslop HE. From specialized centers to community care: expanding the reach of CAR T-cell therapy. Blood 2026;148(15):1855–1857. doi:10.1182/blood.2026035567
38. US Food and Drug Administration. FDA eliminates Risk Evaluation and Mitigation Strategies (REMS) for autologous chimeric antigen receptor CAR T cell immunotherapies. June 2025. Accessed October 2026. https://www.fda.gov/vaccines-blood-biologics/safety-availability-biologics/fda-eliminates-risk-evaluation-and-mitigation-strategies-rems-autologous-chimeric-antigen-receptor
39. US Department of Health and Human Services / Centers for Disease Control and Prevention. Fact sheet: U.S. withdrawal from the World Health Organization. 2026. https://www.cdc.gov/global-health/resources/us-withdrawal-who.html
40. Sadikin BG. Remarks at a WHA79 side event "Stress-Testing Supply Chain Resilience." Geneva, 18 May 2026.

About the Author:

Boro Dropulić is executive director of Caring Cross, a nonprofit organization focused on access to advanced therapies in the public health sector, and chief executive officer of Vector BioMed, a public benefit corporation and viral vector manufacturer that enables companies to deploy CAR-T cell and other advanced therapies in their regions. Reach him at boro.dropulic@caringcross.org.