Autologous cell therapies present a unique paradigm shift in terms of biologics manufacturing from a “one batch for many patients” to “one batch for one patient” model. Over the past two years, FDA approvals of genetically modified cell therapies Kymriah (Novartis) and Yescarta (Kite/Gilead) for haematological cancers have generated tremendous excitement around autologous CAR-T therapies. There are now close to 100 autologous CAR-T/TCR products in clinical trials in the US alone, with around 40 of these in Phase 2 and Phase 3.
Commercialisation of these therapies implies a significant increase in the number of patients to be treated - from hundreds to potentially hundreds of thousands. A reliable manufacturing process that ensures continuity of supply for these last-line treatments, from the point of acquisition of patient tissue through the manufacturing process and final infusion back into the same patient, is critical.
Current standard workflow for autologous CAR-T therapies is often comprised of multiple process steps that are reliant on a mix of open manipulations and limited unit operation level automation. This can present a high risk in terms of process consistency, hygiene and ultimately batch success and product safety, especially when the number of batches to be manufactured increases by orders of magnitude from clinical to commercial. Moreover, since the drug product cannot be sterile filtered, the entire process must be run aseptically. Most importantly, systems that facilitate manufacturing scale-out while meeting all process requirements, as well as controlling cost of goods sold (COGS) to allow reimbursement, are key to the development of this modality.
Single-use technologies (SUTs) are the only practical option to meet the requirements of autologous cell therapy manufacturing and are already exclusively used for this purpose. However, there are a number of challenges with SUTs that need to be overcome as processes move into commercialisation. Firstly, critical consumables (eg. specialised cell culture vessels) are often single-sourced and may pose a high risk to supply chain. Secondly, due to the need for aseptic open or partially open operations, manufacturing normally occurs in a Grade B cleanroom with open processing steps occurring in a Grade A biosafety cabinet (BSC). Consequently, a large number of consumables must be passed through areas of increasing classification while maintaining material sterility. This leads to a high reliance on manual sanitisation, increasing the burden on technicians and also the risk of contamination. Lastly, the impact of leachables, extractables and particulates from the plastics in SUTs on cell growth, performance and drug product quality needs to receive more attention. In light of these challenges, it is critical to optimise the number and volume of single-use consumables that are included in the manufacturing process and at the same time ensure good control over the source and quality of these parts.
Another area of significant risk to commercialisation is from labour-intensive manual aseptic processing. As the number of batches required multiplies, so does the need to recruit, train, qualify and retain a highly motivated manufacturing labour force. The time and effort needed to conduct adequate training and retain proficiency are significant and pose a major bottleneck to scale out. Overall, the large number of single use consumables that must be transferred into cleanrooms, open processing activities and the laborious nature of the processes lead to a high risk profile, as well as high costs which will severely limit access to these curative therapies.
These challenges can only appropriately be addressed through closed, controlled, automated operations with a limited number of integrated SUTs that are appropriately qualified and with good control over the source of supply. Currently, such available automation technologies for autologous cell therapies can be split into two groups: unit operation automation, where each connected step is automated in an independent proprietary machine, and fully integrated systems where most or all of the process is performed in an automated closed system.
Examples include automated units that represent a significant step towards reduced labour and improved process and hygiene control, but still require routine manual intervention introducing variability and contamination risk. Operations must remain within a Grade B cleanroom to enable product transport between equipment as well as multiple sterile welding and disconnect operations. The number of consumables from a variety of sources remains high and this continues to pose a challenge in terms of control.
The next generation of automated systems for autologous cell therapies comprise fully integrated processing capabilities from selection to culture to formulation. Systems such as the Lonza Cocoon and the CliniMACS Prodigy system from Miltenyi are designed to enable automation of most sequential unit operations for a CAR-T process within a single system. An integrated sterile SUT assembly is used for the processing and sampling with built-in process monitoring and control capability. Scale out is achieved through multiple units for parallel operation with each unit dedicated to a single patient process through the duration of manufacturing. The modular nature, high level of traceability, integrated process controls and highly compact form, particularly of the Cocoon, are factors that facilitate effective scale-out and a range of manufacturing options from centralised to point-of-care operations. Such integrated automation systems significantly reduce the number of SUT components required for processing and effectively addressing a number of the challenges described previously. Most importantly, having a validated closed, automated system opens up the possibility of eliminating the need for Grade A and B clean room operations, thus significantly improving therapy COGS.
As the current pipeline matures and more therapies need to scale to meet demand, it is imperative that the industry migrate rapidly to end-to-end automation and other similar single-use manufacturing technologies. Only by fully combining unit operations into an integrated automated system would it become possible to serve the full population of patients needing these therapies and ensure continuity of supply at a reasonable cost.