A hospital team in Bangkok manufactured, tested, and administered genetically modified T cells to patients with aggressive lymphoma in twelve days. This timeframe, documented across twelve patients between 2020 and 2025, is significantly shorter than the delays typically reported for commercial CAR-T therapies in high-income countries. The study, published in Molecular Therapy Oncology in 2026, shows that Bangkok can manufacture CAR-T therapies on-site, and yet this model remains an exception.
The essentials
- A Thai hospital platform produced CAR-T cells locally with a median timeframe of 12 days between cell collection and reinjection. This study provides detailed data on a Thai cohort with B lymphoma, but is not the first demonstration of feasibility in a middle-income country (Molecular Therapy Oncology, 2026).
- Commercial CAR-T therapies cost between 300,000 and 500,000 dollars per patient in the United States; the Thai model aims to reduce this cost through local production, without outsourcing to a centralized manufacturer.
- Documented obstacles are multiple: cost, specialized infrastructure, reimbursement, logistics of centralized manufacturing, and regulation.
- A Jordanian experience published in the same field shows that a similar result is reproducible in another context of constrained resources.
- Between 2030 and 2040, the spread of this model will depend on collective decisions about manufacturing standards, capacity funding, and control mechanisms that allow local production without lowering safety requirements.
CAR-T cells: mechanism and production constraints
CAR-T therapies work on a principle simple to state and difficult to manufacture. White blood cells are extracted from the patient—T lymphocytes, soldiers of the immune system. They are genetically reprogrammed to recognize and destroy cancer cells bearing a specific marker, CD19 in the case of B lymphoma. Then they are reinjected into the patient, now equipped with an instruction that their genome did not contain.
The logistical problem begins at extraction. The cells are alive. They cannot afford to wait. The delay between extraction and reinjection must be managed to the day, and each step—cell culture, viral modification, quality control, freezing—requires equipment that exists in only dozens of sites worldwide. When an American or European patient enters a CAR-T trial, their cells often fly by plane to a specialized factory in Pennsylvania or Switzerland, are processed for two to four weeks, then return, if they return at all.
In the United States, the price or payment for a commercial CAR-T is often around 400,000 dollars or more, and the total cost of care can be higher.
For many patients in these countries, access to international commercial chains remains limited, but local chains already exist in Thailand and Jordan. Centralized international circuits add delays, logistical complexity, and high costs. Commercial CAR-T therapies are largely inaccessible in these contexts, even as the prognosis for refractory B lymphomas is poor.
Bangkok manufactures on-site: the logic of point-of-care
The team at King Chulalongkorn Memorial Hospital chose a different architecture. Rather than sending cells to a centralized manufacturer, it built production capacity inside the hospital. The clean room, a controlled sterile environment, is installed on-site. The clinical lentiviral vector was obtained from Miltenyi Biotec; cell manufacturing was performed locally. Quality control happens before the cells travel even a meter down a corridor.
The result is a median twelve-day delay between extraction and reinjection. Across twelve patients treated between 2020 and 2025, all with relapsed or refractory large B cell lymphoma, nine with de novo DLBCL, two transformed forms, and one primary mediastinal B lymphoma, the team documented initially encouraging response rates. Complete data are available in the Molecular Therapy Oncology article published in 2026.
Twelve patients is a small number. The authors emphasize this themselves: this is a proof of concept, not a phase III clinical trial. But it is precisely what the medical literature lacked: proof that the model works, in a hospital that does not have the resources of a leading American center. This proof of concept opens space that a decade of articles on access inequalities had failed to create.
Jordan confirms that Bangkok is not an exception
The Jordanian experience published in JCO Global Oncology adds data that transforms an isolated case into a signal. A team at the King Hussein Cancer Center in Amman conducted a similar approach, with different but comparable resource constraints. The timeframes and quality requirements documented overlap with those in Bangkok.
Two countries, two continents, two distinct economic contexts, and the same feasibility result. The point-of-care production model is not a Thai curiosity. It could be transferable to specialized centers with regulated infrastructure, quality systems, and experienced teams, but its universal replication is not demonstrated.
Together, the two studies show that scientific competence exists in far more places than the pharmaceutical industry suggests. What is lacking is investment in physical capacity: clean rooms, cryogenic equipment, local logistics circuits, and ongoing team training. This capacity challenge is distinct from the patent challenge and calls for different responses.
The question of funding for healthcare workers in these contexts is also central: as the experience of fragile health systems shows, retaining competent professionals is a prerequisite for any system capable of administering complex therapies.
The current industrial model was not designed for diffusion
Among the anti-CD19 CAR-T therapies used for certain B lymphomas are Kymriah from Novartis, Yescarta from Kite-Gilead, and Breyanzi from Bristol-Myers Squibb, but other CAR-T therapies are also approved depending on indication and jurisdiction. These companies have invested billions in centralized manufacturing sites located in North America and Western Europe.
This model is consistent with their economic logic. But it creates an architecture that makes diffusion structurally difficult. Transportation delays to distant markets add logistical complexity. Prices are calibrated to the payment capacity of insurers in wealthy countries. And dependence on an international logistics chain exposes patients to disruptions, customs delays, freezer breakdowns, air transport interruptions.
The point-of-care model proposes a break with this architecture. It decentralizes manufacturing. It eliminates long-distance transport. It adapts cost to the local context. But it requires initial investment in capacity that most hospitals in middle-income countries do not yet possess.
The political choice is made at this level.
Decisions on the localization of cell therapy production by 2040
The question posed by Bangkok and Amman goes beyond medicine. It concerns the architecture of precision medicine for the coming decades.
Two trajectories are taking shape. In the first, the centralized industrial model remains dominant. CAR-T therapies and their successors—TCR therapies, allogeneic approaches that would use cells from donors rather than the patient themselves—are produced in dozens of global sites and exported to markets able to pay for them. The rest of the world waits, at best, for patents to expire and biological generics to appear, a process that takes a generation for conventional medicines, and whose feasibility for cell therapies remains to be proven.
In the second trajectory, regional hospital networks gradually build local manufacturing capacity, with support from multilateral funding, development banks, climate-health funds, and technology partnerships with manufacturers willing to share protocols. This trajectory already exists in embryo: the WHO launched the mRNA technology transfer program in 2021; in 2023, it concluded an agreement for a global center for bioproduction training, and several Asian governments are investing in hospital biotechnology infrastructure. But it requires collective decisions that no one has yet made at the necessary scale.
Three barriers separate the possible from the probable. First, manufacturing standards: good manufacturing practices defined by the FDA or EMA are often inaccessible without massive investment. Adapted standards for resource-constrained contexts, without lowering safety requirements but calibrating them differently, remain to be built. Bodies like the European Medicines Agency have begun thinking about approval pathways adapted to third countries, but without concrete results to date.
Next, capacity funding. Building a hospital clean room represents significant investment depending on country and applicable standards. Training a team capable of producing CAR-T cells requires long-term commitment. These investments do not fit into the usual models of public health financing, which cover clinical procedures but rarely manufacturing infrastructure. Multilateral funding bodies could play a role, if cell therapies enter their funding scope, which is not assured.
Finally, post-treatment monitoring mechanisms. Administering CAR-T therapy requires monitoring the patient for several weeks to detect complications, notably cytokine release syndrome, a severe inflammatory reaction that can be fatal if not managed rapidly. Management must be assured by a center capable of quickly managing toxicities, with immediate access to tocilizumab; systematic availability of an intensive care team is not established as a universal condition. Without them, the risk of producing locally without being able to manage complications is real. The Thai and Jordanian programs integrate manufacturing, quality controls, and clinical management, without isolated causal proof singling out this whole as the sole explanation for results.
The signals to watch in the coming years are readable. The adoption of simplified protocols by international regulatory bodies, the first technology transfer agreements between manufacturers and hospitals in the Global South, and the emergence of dedicated multilateral funding for cell manufacturing infrastructure would constitute tangible markers of incipient diffusion. Conversely, increased market concentration among a few industrial players and the absence of adapted regulatory frameworks would signal that the point-of-care model will remain a documented curiosity rather than a deployed alternative.
The question of the concentration of advanced medical capacity in a few metropolitan areas is not unique to cell therapies; AI occupations know the same geography, and the answers to be given resemble each other: deliberate investment in peripheral capacity, adapted standards, and political will not to let the geography of access crystallize.
What Bangkok has achieved and remaining challenges
The Thai study does not prove that CAR-T therapies can reach all patients who need them. It suggests that a well-equipped Thai hospital can produce CAR-T therapies with initially encouraging results and indirectly comparable outcomes, without proving it at the level of global reference centers.
It is a proof of concept, not a promise of deployment. The distance between the two remains large. It is measured in infrastructure to build, teams to train, standards to adapt, funding to mobilize, and political decisions that neither hospitals nor researchers can make alone.
Debates on access to medicines have long focused on patents, prices, technology transfer, production capacity, regulation, and skills. The experiences of Bangkok and Amman show that local manufacturing constitutes a lever for access and cost reduction, open to a broader range of actors than expected. Access depends on multiple factors: infrastructure, skills, technologies, inputs, regulation, and market conditions. Hospital infrastructure, training, and institutional investments remain important levers, already mobilized by multilateral programs and existing regulatory frameworks.
By 2030, hospitals that will have invested in these capacities, with or without industry support, will draw the real geography of access to cell therapies. This will be very different from the current map. Or it will not be.
Sources
- Feasibility and outcomes of decentralized point-of-care CD19 CAR T cell for relapsed large B cell lymphoma in resource-limited settings, Molecular Therapy Oncology, 2026
- CAR T-Cell Therapy in Resource-Limited Settings, JCO Global Oncology
- Technology transfer program for biological therapies, World Health Organization (WHO), 2023
- Data on approved CAR-T therapy pricing (Kymriah, Yescarta, Breyanzi), Institute for Clinical and Economic Review (ICER), reports 2022-2024
