Will automation transform the speed and efficiency of cell therapy manufacturing?

Introduction

In 1998, Isabelle Rivière was recruited to lead a new cell and gene therapy facility at the Memorial Sloan Kettering Cancer Center (MSKCC) in New York. MSKCC was pioneering cell therapy for blood cancers, and Rivière, who has a doctorate in cell and molecular biology from MIT and the Curie Institute, Paris, was in charge of developing the manufacturing protocols. “I started with a single technician, and had to build it all up from scratch,” she says.

With Rivière as its director, what became the Michael G. Harris Cell Therapy and Cell Engineering Facility at MSKCC has grown by leaps and bounds. It employs 40 people and occupies 6,000 square feet in the heart of Manhattan. Approximately 500 patients have been treated with the therapies made there. But as the scope of cell therapy has broadened from blood cancers to target solid tumours, and includes more varied treatment modalities, Rivière is finding it difficult to increase production.

Manufacturers of products such as cars and computer chips have moved steadily to adopt automated and digital systems that boost operational efficiencies and lower costs. Companies using them have a better understanding of their processes and costs, and plant operators have an improved ability to make real-time decisions that improve product quality.

In contrast, manufacturers of cell therapies are still figuring out how to best automate and digitize their operations. These are “the first in history to manufacture complex living products,” says Krishnendu Roy, director of the National Science Foundation Engineering Research Center for Cell Manufacturing Technologies (CMaT) at Georgia Tech, in Atlanta. If they can streamline and scale up manufacturing, they will deliver more therapies faster to the patients who need them. But they face unique challenges.

A quest to automate

When Rivière arrived at MSKCC, cell manufacturing protocols were virtually nonexistent, and most steps in the process were dependent on manual operations. The only automated tool was a cell culture system called a wave bioreactor, which rocks gently back and forth so that oxygen mixes better with growth media. Rivière adapted it for the purpose of expanding genetically modified cells in larger numbers.

Over time, Rivière and her team reduced the need for lab workers to handle cells and other materials. For instance, they replaced traditional centrifuges, which required researchers to manually add and remove cells and buffer, with automated cell washers, and they began using sterile connecting devices and tubing to move cells from one step in the process to the next. Over the past five years, Rivière has been validating a fully integrated commercial manufacturing system that streamlines cell processing workflows all the way from cell separation to formulation of a final product.

Rivière now needs digital solutions that reduce manpower costs and the potential for human error. Automated, real-time monitoring of cells, the nutrient broth they grow in, and mechanical parameters such as feed-in rates, “would allow us to determine critical quality attributes for making more consistent products,” Rivière says. Cellular parameters to monitor could include concentration, viability and phenotype, while cell-culture media parameters could include pH, nutrients and metabolites.

This data would feed into an automated control system that responds in real time by adjusting the make-up of the cell-culture media. “If glucose is important for maintaining cell quality, for instance, then you want a system that will replenish glucose as levels fall, without a human having to make that decision,” says Damian Marshall, who leads the assay development and validation team at the Cell and Gene Therapy Catapult, a nonprofit technology innovation center in London, UK. “Basically, what we're talking about is a system that monitors parameters and triggers reactions.”

Defining quality

Any manufacturing operation requires steps to assess the quality of the product, but it’s simpler to assess how well a car or a microchip performs than it is to assess a cell therapy. The clinical benefits of a cell therapy for cancer, for example, can vary with cell type, prior treatment, cancer stage, and a patient’s age. “We’re talking about a very complex interaction between a living entity that’s put into a living patient,” Roy says. “At this stage, we still don't fully understand what is meant by a good-quality cell that predicts optimal functioning in a patient.”

To find out, researchers are looking deeper. Academic and industry partners of Georgia Tech’s CMaT are using transcriptomic, proteomic, metabolomic and other assays — especially at the single-cell level — to identify critical quality attributes, while also identifying the process parameters that control them, Roy says. Once those parameters are identified, cell manufacturers will need new types of sensors, new ways to digitize the data they produce, and new computational tools to process and synthesize vast amounts of data. The ultimate goal is to create bioreactor systems that are more automated than today’s.

The first steps to scaling up

To scale up cell-therapy manufacturing, opportunities for automation need to be identified and acted on as early in the process as possible, experts interviewed for this story agreed. Cell therapies most often go to very sick patients in dire need of new treatment. For that reason, clinical trials of those therapies tend to progress faster than studies of more traditional drugs, and cell therapy manufacturers need to be ready when they’re done.

To avoid delays, facility directors and operators should think early about where automation fits into the manufacturing process, so they can integrate the technology as soon as regulators authorize them to produce and sell the therapy commercially, Marshall says. Because of regulatory requirements for cell therapy manufacturing, he adds, “it becomes more complicated to make process changes once cell therapy products enter phase 3 trials, including the integration of automated technologies [for] scaled-up manufacture.”

Despite the need, opportunities to test and optimize automated processes at smaller scales are lacking today. “Since we generally don’t have scalable instruments, we have to optimize at large scale, which costs a fortune and is often impossible to do because of insufficient patient source material,” Rivière says. “We could progress faster if we could optimize processes in smaller volumes, say 10 milliliters, and then scale up from there. That would require scalable equipment with dynamic ranges, combined with real-time, online monitoring of key process parameters," she says.

As cell therapy matures as a field, MSKCC is developing cell therapies for a wider variety of diseases, and clinical trials conducted there are examining more combination treatments than ever. This makes automation a critical need, Rivière says. “In my view, it is not possible to expand applications and access without automation and digital monitoring of the operation.”

Cytiva supports translational researchers from the early phases of preclinical work through clinical trials and commercialized products with an array of cell and gene therapy manufacturing instruments, process development services, and support solutions.