CAR T cell therapies have revolutionized treatment options for patients with certain blood cancers, but persistent challenges remain—among them, variable patient responses and relapses that occur even after an initially strong therapeutic effect. A new study led by Mohamed Abou-el-Enein, MD, PhD, at the University of Southern California, aims to bring new precision to the field by offering a much more granular understanding of the CAR T cells being delivered to patients.
In a recent Molecular Therapy paper, Abou-el-Enein’s team details the development of a high-dimensional flow cytometry panel that captures a detailed portrait of CAR T cell phenotype across the manufacturing timeline. Rather than looking at just a handful of surface markers, this new method simultaneously assesses 36 characteristics related to cell function, exhaustion, proliferation, and differentiation—all on a single-cell level.
“This project was driven by a need to better understand the biology of the CAR T cells we’re actually infusing into patients,” Abou-el-Enein said. “Although we have standard quality control testing at the end of manufacturing, it gives us a very limited view. We wanted to go much deeper.”
The method relies on spectral flow cytometry, a powerful platform capable of analyzing dozens of markers on individual cells. Abou-el-Enein likens it to getting a full medical checkup, rather than a single lab result.
“You can see how each of the markers interacts and evolves during manufacturing,” he said. “It’s technically complex and time-consuming, but the level of information you get is unprecedented.” Cells are passed through the cytometer, where lasers cause the fluorescent tags to emit light that can then be detected and measured. This indicates whether a specific molecule is present and how strongly it is expressed. The spectral flow cytometer provides a more comprehensive picture of each cell compared to standard tools, which can only measure about 10 markers at a time.
A significant focus of the study was how CAR T cells evolve between day 5 and day 10 of ex vivo expansion—a standard window used by clinical manufacturing teams. The team found that by day 10, the cultures contained a significantly higher proportion of terminally differentiated effector T cells. These cells are primed for cytotoxic activity but have limited persistence in the body. In contrast, day 5 products retained more memory-like T cells, particularly stem cell memory T cells, which have long-term proliferative potential and can offer more durable responses.
“For example, the stem cell memory population was about 25% at day five and dropped to 11% by day 10,” Abou-el-Enein noted. “On the other hand, the fully differentiated effectors doubled in frequency by day 10.”
Despite these shifts, both day 5 and day 10 cells were equally effective at killing target tumor cells in vitro, highlighting that functional capacity alone doesn’t capture the complete therapeutic picture. This kind of insight opens the door for more tailored manufacturing decisions.
The 36-marker panel captures not only cell phenotype but also functional features, activation status, metabolic readiness, and exhaustion levels—all at multiple time points. Previously, most manufacturing pipelines included only a final product test, often using a limited panel of 4–6 markers. That approach, Abou-el-Enein emphasized, provides just a snapshot.
“CAR T cell therapy is still a very young field, and we’re only beginning to appreciate how dynamic these products are during manufacturing,” he said. “We need better tools to track these changes in a detailed, standardized way. This work fills a critical gap in our understanding of how manufacturing conditions shape the therapeutic potential of CAR T cells. By pinpointing when CAR T cells acquire—or lose—functional fitness, we can now tailor the timing of cell manufacturing, which could immediately impact clinical decision-making and patient outcomes.”
While this study focused on manufacturing-phase analysis, Abou-el-Enein’s lab is already looking ahead to applying the panel directly to patient samples.
“If we can track how the product changes in vivo, and link those changes to clinical outcomes, we’ll be much closer to personalizing CAR T therapies. Right now, we don’t have that level of granularity,” he said. “This approach could help change that.”
