Source: Comprehensive Cell Solutions, A Business Unit of NYBC Enterprises
The complexities inherent in today’s biopharmaceutical development paradigm have necessitated novel approaches for early phase workflows that can accelerate timelines and reduce cost and risk. For both process and analytical development (PD and AD) in the advanced therapy space, this drive toward more automated, closed processing has led to unprecedented innovation aimed at standardizing these workflows to the same degree as other, more incumbent therapeutic modalities.
Yet this pursuit of fully automated, completely closed PD and AD can serve to create new challenges for a workflow – going “all in” on a platform has the potential to introduce inflexibility, limit scalability, and complicate tech transfer. This is largely due to the diversity and complexity of the CGT products in the current development pipeline, which often require customized approaches and bespoke solutions to address the unique variables that accompany their optimization.
There are a number of technological solutions that, when applied to a cell and gene therapy workflow, can streamline development and improve workflow integration. By pursuing targeted automation and modularity for a program, developers can eliminate common workflow bottlenecks, improve manufacturing reproducibility, and establish a highly flexible and controlled development paradigm that can adapt to technical and business considerations.
Striking the Balance: Modular vs. Fully Automated Development
When it comes to both PD and AD for advanced therapies, the drive toward optimizing processes early has led to an increased focus on automation. In the context of advanced therapy production, this automation is likely to look much different than that which typifies small molecule production. Moreover, the diversity of products in the cell and gene therapy pipeline makes standardizing development and manufacturing particularly challenging. Even the most well-characterized modalities, like mAbs, require customization in order to balance reproducibility with the need for manual interventions. This balancing act can be further complicated by analytical considerations – an organization’s understanding of its assets is often not comprehensive enough in the earliest stages of development to inform an optimal late-stage analytics strategy.
Ultimately, one of the key variables impacting scalability and commercial feasibility for automated processes is this focus on downstream considerations. The realities of advanced therapy development and production can make a fully closed, fully automated system a doubled-edged sword, as it precludes the possibility of reevaluating or revamping a discrete process step without revisiting an entire workflow.
Modular platforms that incorporate targeted automation for the process steps best suited to these approaches can create greater flexibility as an application moves through phases, allowing operators to make adjustments to discrete steps without impacting other parts of the process.
Whether a process is fully manual, fully automated, or somewhere in between, much of the analytics supporting it are likely to be performed offline. This has made analytics distinct from a process; integrating analytics through the adoption of more in-line and online process analytical technologies (PAT) that incorporate greater degrees of automation is therefore one of the central ways developers can realize process improvements, accelerate timelines, and control costs. Advancements have been made in automating analytical methods, such as flow cytometry, ELISA, or cell counting; further advancement of integrating these automated analytics more consistently throughout a process can shave significant time off QC release testing, as well as afford operators the necessary insights to inform near-real time decision-making.
Developers often work to define critical quality attributes (CQAs), critical process parameters (CPPs), and even a therapeutic’s mechanism of action during early development. For many working to develop a cell or gene therapy, a lack of process or product knowledge at the earliest stages of development can make introducing automation difficult, particularly more comprehensive automation solutions that are intended to serve a final manufacturing process.
While modular systems also require this knowledge, their very nature enables operators to incorporate such automation as discrete parts of a process, as products become better characterized. The sampling component of each approach can also greatly impact this characterization – for fully closed, maximally automated systems, collecting in-process samples, especially those that are non-routine, can be challenging or infeasible. Because modular systems allow for both sampling of an individually automated step and the flexibility to retain hands-on steps as needed, developers can ensure access to better, more tailored data from deeper process and product characterization studies, which in turn can enable them to make better decisions around future automation. It can also aid in demonstrating comparability for regulators by enabling more incremental process changes, rather than the major process change transitioning to a fully automated platform would require.
A Use Case: Incorporating Automation to Achieve Optimal Selection
Incorporating automation in an advanced therapy development workflow can hugely streamline nearly every step. In a typical selection process, for example, apheresis from a patient is received, cells are washed, and operators perform immunoprecipitation, often via magnetic beads, before performing separation. An analog version of this process is highly manual: during cell washing, operators must aliquot the product into smaller bags or vials, add media to each portion, perform centrifugation, and then extract the supernatant in discrete, open steps. This is likewise true for the next step – during bead incubation, operators add media and a bead/antibody mixture, induce incubation, add wash media, perform centrifugation, and extract the supernatant.
While this approach has the benefit of being highly flexible in terms of both processing and sampling, the number of manual manipulations required, coupled with operator-to-operator variability, introduces the potential for contamination, batch inconsistencies, and operator error. It also requires more stringent cleaning and isolation procedures, sometimes necessitating ISO 5/7 standards for a facility. This, coupled with the need for more personnel with higher expertise, more equipment, and more facility footprint, and the pitfalls of manual processes compound as therapeutic demand increases.
In contrast, a more modular approach that incorporates a cell washer, to perform volume reduction or buffer exchange by automating the manual steps of analog version, has the potential to streamline these critical selection steps. Since most cell washers and separators can be utilized as part of a closed system, the risk of contamination is lessened; moreover, closing these steps enables operators to easily retrieve sterile, closed samples for in-process testing. This level of automation requires fewer operators, fewer pieces of equipment, and often, less space. It also typically results in a lower ISO standard for a facility, and can enable greater concurrent processing capacity, as operators can have multiple sets of equipment running multiple patient samples simultaneously in the same amount of space and with fewer personnel. This approach can have drawbacks – it still requires a certain number of operators for setup/teardown, monitoring, sample collection, and transitional steps, and training on disparate equipment and different monitoring systems can introduce the potential for error – but it retains flexibility in modifying individual equipment or process steps.
For some companies, the ultimate goal of any automation initiative is to end up at a fully automated process. In this scenario, samples received from patients are introduced to a fully closed system wherein wash, incubation, and separation are all performed within a single, specialized piece of equipment. The potential benefits of such an approach are significant: operator variability is further reduced when compared to modular systems, personnel training is confined to one major piece of equipment, and lower ISO requirements are similar to those needed for modular approaches.
While these advantages may be valuable, they are counterbalanced by a number of possible downsides to a fully automated approach. These include a rigidity that may preclude customization for different programs or cell types, and which may limit the products or modalities suited to the platform. The all-or-nothing nature of full automation likewise can create the need for more capital investment in order to enable operators to study multiple conditions in parallel; similarly, this investment in a single piece of equipment can create significant bottlenecks whenever problems arise. Troubleshooting, both for equipment and for processes, is also complicated, as is sampling while a process is ongoing.
Maximizing Benefit and Minimizing Risk for Cell and Gene Therapy PD and AD
As developers uncover new knowledge about a product and must adjust their processes accordingly, retaining flexibility through modularity serves to improve troubleshooting and bolster process insight. Right-sizing automation to fit a unique product requires evaluating a process and prioritizing certain steps – steps at high risk for contamination, for example, may benefit from automation, as can steps where frequent sampling is necessary. Maximizing value-added changes to a PD or AD paradigm is core to introducing the right degree of automation; balancing these gains with potential drawbacks stemming from the rigidity that can accompany full automation can help organizations arrive at the right fit for their purpose.
Modular automation allows operators to close process steps wherever possible while retaining flexibility and enabling greater redundancy for many parts of a process. Coupled with analytical solutions that can be integrated and connected to inform faster decision making, this approach to automation is well-suited to the complexities that accompany contemporary advanced therapy development. For companies manufacturing autologous therapies or managing a diverse development pipeline, the need to transition a process from product to product or sample to sample as seamlessly and efficiently as possible necessitates technology solutions aimed at creating flexibility.
Ultimately, balancing this need for greater speed and lower risk with the realities of manufacturing highly bespoke therapies requires systems that can evolve alongside an asset during development and through commercialization.