From centralized to decentralized

Cell and gene therapies cover a variety of products and approaches, which are principally categorized as cell-based and non-cell-based products. Non-cell-based gene therapy products are similar to biotech products in terms of manufacturing and stability and can be distributed frozen/freeze-dried without any particular critical issues.

Cell-based products, on the other hand, have significant limitations in terms of shelf life, both for the biological starting material (blood or other tissues) and for the finished product. In general, to achieve sufficient numbers of cells during the manufacturing process to guarantee the dose and potency, it is preferable to work with fresh tissues and cells.

Freezing the starting material greatly facilitates transport from procurement sites to the manufacturing site. However, this impacts the viability of the starting tissue and, therefore, the yield of the manufacturing process. For some cell-based products, the starting material is also very scarce due to patients’ health conditions and treatments with medications that cause cell distress, such as chemotherapy drugs, so it is preferable not to further stress the cells by freezing.

For cell-based drug products consisting of viable cells as drug substance, potency is closely related to the viability of the cells themselves. The number of cells in the drug product decreases steadily before administration, and consequently so does its potency. This occurs because, in accordance with clinical standards, the final formulation must meet stringent quality and safety requirements for patient administration and, as such, cannot function as a nutrient medium for cells. Formulated cells therefore begin to starve and die before administration, resulting in a very limited product shelf life.

While many large hospitals have in-house manufacturing units for cell and gene products and can work with fresh tissue and administer fresh drug products to the patients, the distribution of these products remains limited. As a direct consequence, despite seven approved autologous CAR-T therapies in the U.S., only 20% of eligible patients who could benefit from treatment can access it.¹

However, having several production units for the same product within the points of care (POCs) distributed consistently across the territory would allow more patients to be treated. This decentralized model involves establishing a central site that communicates with regulatory authorities and manages decentralized manufacturing units (DMUs) at POCs. Depending on local regulations, the responsibilities of the central site and DMUs may differ, but the same principles are generally applicable.

Benefits of producing at/near POC

The proximity of the cell and gene manufacturing unit to the patient’s bed allows for working with fresh tissues, administering fresh drug products, reducing overall costs and shortening treatment times, enabling the timely treatment of patients with rapidly progressing diseases (e.g., pediatric oncology patients in critical condition).

Avoiding freezing of drug product means not only increasing their potency and, consequently, the efficacy of the treatment, but also improving their safety profile, as cryoprotectants such as dimethyl sulfoxide (DMSO) do not need to be used as excipients. Adverse events such as nausea, vomiting, and abdominal cramps occur in up to 35–50% of patients during the infusion of DMSO-containing products. Dyspnea and hypoxia can occur, especially in patients with compromised pulmonary function.²

Furthermore, the starting material can also be processed immediately, increasing the probability of achieving the dose even in cases where there are limitations on the amount of blood/tissue that can be collected.

Closeness to patients also cuts out complex logistics for transferring the tissue to be processed to the production site and returning the drug product. Supply chain failures may lead to product degradation or contamination, resulting in the cancellation of a patient’s treatment when no therapeutic alternatives are available. A simplified supply chain also has an impact on cost reduction, limiting the need for shipments and the storage of refrigerated or cryopreserved biological starting materials and drug products.

Challenges of producing at/near POC

When manufacturing sites for the same drug product are decentralized within POCs, it’s the responsibility of the central site to guarantee that all the DMUs met the required GMP standards. To reduce the batch-to-batch and site-to-site variability, it is essential to standardize the manufacturing process and the analytical methods, ensuring reproducibility of the manufacturing process at each DMU. 

Avoiding freezing of drug product means not only increasing their potency and, consequently, the efficacy of the treatment, but also improving their safety profile.

Automated or robotic systems certainly help to ensure reproducibility between different DMUs. Manual processes should be avoided to reduce the variability associated with manual labor, making AI-driven robotic systems preferable. Validating these new systems poses new challenges, but regulatory agencies’ efforts to publish considerations and guidelines suggest this is the right direction to take.

Furthermore, all cell-based manufacturing process steps must be conducted under aseptic conditions, because cells cannot be terminally sterilized by filtration or other means, but must remain viable. Closed systems, such as isolators or truly closed automated systems, are necessary to ensure asepsis and can be maintained more easily than an open process in a grade B clean room.

The ancillary materials, critical raw materials, consumables, and reagents used in the process must meet GMP-like standards and be consistently available. This aspect is often underestimated, but it can become a blocking factor for production in some geographical areas. A robust risk assessment that also accounts for the procurement of all materials necessary for production and quality controls must be carried out in the preliminary stages of process development.

Another critical aspect is knowledge of the manufacturing process at the DMUs. The involved teams of the DMUs must be highly specialized and must receive proper training from the central site team, to avoid risk of human error or operational deviations during clinical batch manufacturing. The manufacturing process is, in fact, commonly developed at the central site and then transferred to the DMUs. At the early stages, when there is little historical data, a direct line of communication between the central site’s R&D and the DMU is essential to avoid failures.

Finally, the role of the central qualified person (QP) or quality assurance (QA) in batch release requires particular attention. The release of these types of products is generally carried out in two steps: first, the product is released for administration, based on partial QC data, including essential safety and identity tests as a minimum; then, once all QC data are available, the batch is definitively released.

The central QP/QA must have an efficient tool, such as ad hoc software to supervise decentralized production in real time to speed up the release of the batch for administration to the patient, which is very time-sensitive due to its short shelf life.

It is also essential to build trust between the DMU staff and the central QP/QA. An auditing plan for each DMU, and the central-site DMU, plays a crucial role in creating an efficient team. Here, it is highly trained people and their ability to cooperate for a common goal that makes the difference.

In conclusion, decentralized manufacturing can improve access to innovative medicines for unmet medical needs by reducing the time for the clinical availability of the treatment and the associated costs. However, to achieve this goal, it is important that the central site, decentralized manufacturing units, and point of care work as a single, close-knit team with each patient’s health as their sole focus. 

References

1. The USD 30k CAR-T therapy: a future within reach? (2025). KPMG.
2. Hypersensitivity reaction to dimethyl sulfoxide. (2023, Feb). Cancer Institute NSW.