What you will learn in this 31-minute webinar:
- >> How real-time, label-free tracking with the Axion BioSystems Maestro Z enables continuous monitoring of tumor killing dynamics, allowing for the evaluation of different CAR T cell engineering strategies
- >> The role of maintaining a stem-like phenotype in nonactivated CAR T cells can lead to enhanced persistence, metabolic fitness, and long-term effectiveness
- >> Improve CAR T therapeutic efficacy through metabolic reprogramming strategies
- >> Enhancing CAR T survival and function in the TME with cytokine expression and metabolic fitness optimization
About the presenter:
Saba Ghassemi, PhD
Saba Ghassemi, PhD, is an Assistant Professor in the Department of Pathology and Laboratory Medicine and serves as a Principal Investigator at the Center for Cellular Immunotherapies at the University of Pennsylvania. Her research focuses on developing novel CAR T cells with enhanced potency through a multidisciplinary approach that integrates engineering with CAR T cell immunology.
Dr. Ghassemi pioneered the development of an abbreviated culture paradigm, resulting in less differentiated progeny with improved potency. This breakthrough led to a successful clinical trial at UPenn utilizing a 3-day manufacturing process. Additionally, she has devised a novel method for delivering CAR transgenes into quiescent T cells, enabling the generation of highly potent CAR T cells within 24 hours. Dr. Ghassemi is currently spearheading efforts to transition this innovative platform into an upcoming clinical trial at Penn.
Her groundbreaking work has garnered recognition through several research grants and patent applications aimed at enhancing CAR T cells' efficacy, expansion, and fitness for adoptive immunotherapy. Dr. Ghassemi's ongoing research endeavors focus on optimizing, streamlining, and automating the manufacturing process of CAR T cells, with the overarching goal of enhancing the translational applicability and accessibility of these novel therapies to a broader range of geographical locations and patient populations.
Transcript of the webinar:
Charlotte Barker:
Hello, and a very warm welcome to today's Cell and Gene Therapy Insights webinar, titled “Armored CAR T Cells with Enhanced Potency for Cancer Immunotherapy.”
I'm Charlotte Barker, an editor at BioInsights, and joining me today is Dr. Saba Ghassemi, who will explore how maintaining a stem-like, non-activated CAR T cell phenotype—and using metabolic and cytokine engineering strategies—can enhance CAR T persistence, function, and long-term anti-tumor efficacy.
After the presentation, we’ll have a live Q&A session. We invite our audience to pose their questions to Dr. Ghassemi using the “Ask a Question” box at the bottom of your screen, and we’ll try to get through as many of those as we can during the session.
I’d also like to draw your attention to the resources on the right-hand side of the screen, where you can find more information on the topics covered today.
Now, I’ll just briefly introduce our presenter: Dr. Saba Ghassemi is a leading researcher at the University of Pennsylvania, where she develops potent, rapidly manufactured CAR T cells through innovative engineering and immunology approaches to enhance their clinical efficacy and accessibility.
So, without further ado, I’m going to hand over to Saba to kick us off with the presentation.
Dr. Saba Ghassemi:
Thank you very much for the introduction. I would like to talk about the advances that we’ve made over the past couple of years on the non-activated CAR T cell platform.
As many people are familiar with from this slide, this is an overview of the process of generating CAR T cells for adoptive immunotherapy. T cells are isolated from the patient and cultured ex vivo over a period of typically 9 to 15 days in a very nutrient-rich environment. They are then genetically modified with a CAR (chimeric antigen receptor).
This is the area I’m most interested in—the highlighted portion—and the overall goal is to increase the potency of our product: the CAR T cells. We have good reason to believe that the current process is suboptimal.
The success of CAR T therapy depends on several parameters, many of which intersect and influence one another. For example, using different co-stimulatory domains in the CAR design typically impacts the persistence of CAR T cells. What we are primarily interested in is improving the quality of our product—CAR T cells.
In general, when I mention "quality," I’m referring to the differentiation state of these CAR T cells and their metabolic fitness. We recently showed that activated T cells progressively differentiate over time. By the end of their expansion period—as shown in the panel on the right, on Day 9—they accumulate more terminally differentiated effector progeny. These cells may have strong cytotoxic activity, but they lack persistence. So, while they’re great killers, they don’t stick around, which ultimately makes them less effective for T cell–based therapies. You can also see in the panel on the left some of the important subsets—such as naïve-like T cells or stem cell memory T cells. After several days of activation and expansion, we completely lose that subset. At this point, many studies have shown a strong correlation between the presence of these less-differentiated subsets in the final CAR T cell product and improved patient outcomes following infusion.
With the success of CAR T therapy, we wanted to further investigate how the duration of cell expansion impacts therapeutic efficacy. In a NSG mouse model of leukemia, we previously showed that after establishing the xenograft in the mice, we infused CAR T cells that had been expanded for varying lengths of time—short (3 days), medium (5 days), and long (9 days).
As you can see in panel C of the graph, the Day 3-expanded cells significantly outperformed the Day 9-expanded cells in terms of anti-leukemic activity.
For this experiment, we intentionally used a subtherapeutic dose to stress the system—meaning, we didn't want the treatment to completely cure the mice upfront. This allowed us to observe meaningful differences in performance. Despite this, the bioluminescent imaging (BLI) measurements clearly showed that the Day 3-expanded cells were able to clear the tumor. In contrast, the Day 9-expanded cells completely lost control over tumor growth.
Additionally, peripheral blood data supported these findings. We observed a much higher number of CAR T cells in the blood of mice that received Day 3-expanded cells compared to those that received Day 9-expanded cells. This increase in circulating CAR T cells strongly correlates with their persistence over time. Building on this insight, a clinical trial was initiated at Penn to test the efficacy of a 3-day CAR T cell product versus the standard 9-day product. This was the first-in-human study to evaluate a humanized CAR-19 construct that was further engineered to secrete the pro-inflammatory cytokine interleukin-18 (IL-18).
Based on the clinical data shown here—and with the study now concluded—we observed an 81% complete response rate, highlighting the superior potency of the 3-day expanded cells compared to the traditional 9-day expansion process. One of the most important aspects of this study was the dose used. Compared to the standard Kymriah process (FDA-approved a few years ago at Penn), which typically uses 600 to 800 million CAR-positive T cells, this new process used only 3 to 7 million CAR-positive T cells. The key takeaway here is that dose alone is not the critical factor—rather, it's the quality of the CAR T cells that matters most. So technically, if you have a lower dose of high-quality T cells, that could be sufficient—possibly even more effective—than a large dose of 600 to 800 million mediocre CAR T cells. This highlights just how important it is to focus on cell quality, especially when using novel platforms that aim to shorten the manufacturing process.
Based on this data, we hypothesized that eliminating TCR activation—specifically the use of anti-CD3 and CD28-coated beads during T cell stimulation—could improve outcomes. If we can still successfully perform gene transfer into the T cells without activating them, this process might yield a T cell product with high functional potency. In other words, we hypothesized that we could generate CAR T cells composed of less differentiated phenotypes—and that adoptive transfer of these less differentiated CAR T cells would result in better persistence and efficacy would have better in vivo functional activity.
So, in this experiment, we again used the same NSG mouse model of leukemia. We compared two different CAR T cell manufacturing processes: The standard process—a Day 9 protocol where cells are activated and expanded over time. In this method, we:
- Isolate the cells
- Activate them using anti-CD3/CD28-coated beads
- Perform lentiviral gene transfer
- Expand them over 9 to 10 days
- And finally, wash, freeze, or inject them into the mice
The non-activated platform—where we:
- Collect the T cells
- Directly mix them with lentivirus without activation
- After 24 hours, we wash and immediately inject the cells into the mice
We used a functional stress test in vivo, administering a limited number of non-activated CAR T cells, and compared the outcomes with those of the standard process. Looking at the graph, you can see that in the Day 9 standard platform, the cells clear the tumor very quickly but all of our mice relapsed in the standard Day 9 group.
In contrast, when you look at the non-activated groups, even at different doses, there was sustained control of leukemia growth. This control was strongly associated with persistence and engraftment of CAR T cells, as shown in the graphs at the bottom of the slide. These results led us to consider: "Could we extend our clinical trial using the non-activated CAR T platform?"
So, in this next phase, we developed two humanized CAR-19 products:
- One with a GFP control, and
- One engineered to express IL-18.
In collaboration with Axion BioSystems, we used their Maestro Z impedance platform to evaluate the real-time cytotoxic output of these cells. This device works by generating impedance when tumor cells are seeded on the plate. As CAR T cells kill tumor cells, impedance decreases—giving us a continuous, real-time readout of cytotoxicity.
In this case, we tested:
- Non-activated CAR-19 GFP, and
- Non-activated CAR-19 IL-18.
As the CAR T cells began killing the target cells, we observed a corresponding drop in impedance, reflecting their cytotoxic activity. One of the major benefits of using this impedance-based system is that it allows for real-time monitoring of killing dynamics—unlike traditional chromium-based killing assays, which only provide data at a single endpoint.
Additionally, chromium assays can be cumbersome and less user-friendly. The impedance platform, by contrast, offers valuable insights into both the potency and kinetics of CAR T cell performance.
As you can see in the graph, in the tumor alone or non-transduced T cell (NTD) groups, the impedance signal increases over time, which reflects tumor growth. However, after adding the effector cells, the impedance signal decreases, and the rate of decline varies by group. The red line represents non-activated CAR-19 IL-18, and the blue line represents non-activated CAR-19 GFP.
To further validate these findings, we next tested these products in vivo using a JO1 mouse model. In the JO1 mouse model, we used the same CAR T products that were tested in the clinical trial—specifically, the 3-day manufactured, activated CAR-19 IL-18.
As a control, we also included activated CAR-19 alone, and compared these with the non-activated CAR T cell platform.
In the upper graph, you can see comparisons both with and without IL-18 expression. The clear winner here was the non-activated CAR-19 IL-18 group—it cleared the tumor effectively and persisted over time.
It’s important to note that all of these experiments used very low doses, intentionally near sub-therapeutic levels, to better distinguish which group could truly outperform the others. To control for any differences due to time in culture—for example, in the previous slide, one group was 3-day activated, and the other was 1-day non-activated—we ran a direct comparison using equal time points. In this setup, both groups were expanded for just one day—one using the activated system, and the other using the non-activated system. We then compared humanized CAR-19 IL-18 cells under both conditions.
In this experiment, we had two groups:
- In one group, we activated the cells.
- In the other group, we did not activate—we only exposed the T cells to lentivirus for gene transfer.
We then compared their performance in vivo. As you can see in the graph, the red line represents non-activated CAR-19 IL-18, and the blue line is the 1-day activated CAR-19 IL-18. Both groups were able to clear the tumor, but the non-activated, humanized CAR-19 IL-18 group outperformed all others. Looking at T cell counts in peripheral blood, by Day 38, the non-activated group showed nearly a full log increase in T cell numbers. This strongly correlates with enhanced persistence of these T cells in vivo.
We also wanted to evaluate their efficacy in a solid tumor environment. For this, we used a pancreatic cancer xenograft model, testing Meso-CAR T cells either expressing GFP or IL-18.
So we had two groups:
- Meso-GFP, and
- Meso-IL-18.
We compared the efficacy of these CAR T cells in the xenograft model using multiple dosing levels to evaluate dose responsiveness. But the real takeaway The main point of this experiment was to determine whether the non-activated CAR T cell platform could also show efficacy in a solid tumor environment. As you can see in the results, these cells do clear the tumor, and their performance is clearly dose-dependent.
For example:
- The red line represents the tumor-only group (no treatment).
- The green and brown lines represent different doses of Meso-CAR (either alone or with IL-18).
As the dose decreases, it takes longer for the CAR T cells to clear the tumor. However, all the groups eventually eliminate the tumor, which strongly supports the potency and quality of these non-activated CAR T cells in terms of tumor clearance. We also wanted to test how long these non-activated CAR T cells could persist in the solid tumor environment.
In this study, we followed the mice for up to 90 days. As shown in the graph, Meso-CAR IL-18 cleared the tumor slightly faster than Meso-CAR alone, even though both groups were non-activated. The key takeaway here is that the non-activated CAR T platform—whether expressing IL-18 or not, and whether in hematological or solid tumor models—shows strong persistence and anti-tumor activity. In a blood-based tumor xenograft model, the non-activated CAR T cells also proved to be very potent.
It’s highly feasible for us to generate CAR T cells within just 24 hours using this non-activated approach. These cells are not only extremely potent, but they also exhibit durable anti-tumor functionality—persisting effectively over extended periods of time. This clearly demonstrates the advantages of this platform compared to the conventional CAR T manufacturing process, which typically involves T cell activation and expansion over two weeks.
Our findings strongly suggest that this longer expansion period is not necessary—in fact, it may be detrimental, as it results in the loss of CAR T cell quality over time. In another experiment, we wanted to determine whether this non-activated platform could be applied using actual patient samples. So, we obtained apheresis samples from patients and applied the non-activated CAR-T manufacturing method. From these samples, we generated two types of CAR T cells:
- CAR-19 GFP (as a control)
- CAR-19 IL-18
We then expanded them over time, and as you'll see in the next section, they maintained functionality and quality, even when derived from patient material. We found that the non-activated CAR-19 IL-18 cells could completely clear the tumor. When we analyzed the T-cell counts in peripheral blood of the treated mice, we observed that the group receiving CAR-19 IL-18 (C19-IL18) had approximately one log more T-cells compared to the control group, which received CAR-19 co-expressing GFP (C19-GFP).
This clearly highlights the advantage of including IL-18, which seems to enhance T-cell persistence and overall therapeutic potential.
These results have given us the confidence to proceed with our next clinical trial. We're planning to extend the CAR-19 IL-18 product—previously manufactured with a 3-day activated process—to a 1-day non-activated CI (clinical implementation) platform. Overall, I hope the data I’ve presented today help convince the field that we’re getting closer to a true point-of-care solution for CAR T therapy.
By using this non-activated, rapid manufacturing platform, we’re not only retaining T-cell quality, but we’re also opening the door to faster, more accessible, and potentially more effective therapies. All the conventional processes we currently rely on are cumbersome, resource-intensive, and often depend on materials that are not easily accessible—for example, human serum. At some point, we may run out of available human serum, or face shortages in other critical reagents needed to expand T cells over two weeks. This prolonged culture also increases costs and the likelihood of human error during manufacturing.
Now, if you consider a 1-day process, it becomes much easier to automate, compared to a 2-week protocol. And automation is another highly active area of innovation in the field. By simplifying the process to 24 hours, we enhance scalability, reduce cost, and most importantly, increase access to CAR T therapy for more patients.
In conclusion, what I hope I’ve convinced you of today is that we are now able to generate CAR T cells within 24 hours using a non-activated platform. These CAR T cells are extremely potent, show enhanced durability, and have superior persistence in vivo when compared to traditional activated CAR T cells. This approach not only simplifies manufacturing but could also bring us closer to a true point-of-care solution for CAR T therapy.
Another important point I'd like to mention is about the cytolysis data generated using the Maestro Z system from Axion Biosystems. This system has been incredibly valuable, as the cytotoxicity data it provides in real-time closely correlates with our in vivo outcomes. What I personally appreciate about it is its ability to give us dynamic, continuous insight into which CAR T cells exhibit faster killing kinetics compared to others.
Additionally, I want to emphasize again that very low doses of CAR T cells are sufficient—as long as they are of high quality. This really highlights the principle: Quality matters more than quantity. So rather than focusing on producing hundreds of millions of low-quality CAR T cells, we can achieve superior outcomes with a smaller number of highly potent cells. for a long time. Unfortunately, during that waiting period, we often lose patients. That’s why it’s so important to develop a rapid, reliable, and accessible CAR T manufacturing process, so we can bring this potentially life-saving therapy to patients faster.
With that, I want to thank all of my collaborators. We truly believe that "teamwork makes the dream work." This work has only been possible through the dedication and coordination of so many amazing individuals, and through our ongoing collaboration with Axion BioSystems—who have been instrumental, not only in providing cutting-edge instrumentation but also in offering valuable feedback on our experimental design. Thank you very much.
That concludes my presentation.
Charlotte Barker:
We’re now seeing lots of great questions coming in, so we’ll go ahead and move into the Q&A session.
First question: Could you describe the rationale for the IL-18 armor strategy? And if IL-18 is expressed, would you define a T cell that expresses IL-18 as activated?
Dr. Saba Ghassemi:
Great question. So, the rationale for using IL-18 really fits into a broader effort we have to enhance the overall quality and functionality of our CAR T product.
We’ve been exploring several strategies—whether through optimization of the manufacturing process, preservation of T cell phenotype, or through genetic engineering—to improve therapeutic performance.
IL-18 is a pro-inflammatory cytokine, and there's strong evidence from prior studies that co-expression of IL-18 enhances CAR T cell function. It helps boost the production of interferon gamma, and overall, improves the anti-tumor activity of the CAR T cells, particularly in challenging tumor microenvironments.
So, by expressing IL-18, the cells become better equipped to “defend themselves” against tumor suppression mechanisms.
As for whether IL-18 expression makes them “activated”—it’s important to clarify that these cells are not activated through TCR (T-cell receptor) signaling in this context. Their activation occurs via CAR engagement with the target antigen in vivo, not from traditional TCR stimulation.
So, IL-18 expression itself doesn’t necessarily mean the cell is in an activated state; it's more about priming the cells for enhanced functional response once they encounter tumor cells.
Charlotte Barker:
In your one-day manufacturing using lentiviral vectors, do you formulate the dose based on transduced T cells or total T cells? And what's the transduction efficiency for one-day versus nine-day manufacturing?
Dr. Saba Ghassemi:
That’s a great question. For the 24-hour manufacturing, we formulate the dose based on total T cells, mainly because at that early time point, CAR expression isn’t reliably detectable by flow cytometry.
That said, we have used PCR-based methods to estimate CAR positivity within 24 hours, and typically we see around 10% CAR positivity. We've been actively optimizing this, and our latest approaches are showing significantly improved transduction efficiencies.
We also correlate these results with CAR T cell counts from peripheral blood in vivo to refine our estimations over time. But again, for initial dosing—especially in animal studies—we rely on total T cell counts.
Charlotte Barker:
Do you use cytokines during your one-day, non-activated T cell manufacturing?
Dr. Saba Ghassemi:
Yes, we do. We use IL-7 and IL-15 during the process. These cytokines are important for T cell survival and homeostatic proliferation, and they help maintain a less differentiated T cell phenotype, which is beneficial for persistence and function post-infusion.
Charlotte Barker:
Can the Maestro Z assay be used to model hypoxic conditions like those found in the tumor microenvironment? And have you tested CAR T cell function in hypoxia using this platform?
Dr. Saba Ghassemi:
Yes, absolutely. The Maestro Z platform is very well-suited for modeling hypoxic conditions, and we’ve had great success using it in that context.
We've metabolically engineered CAR T cells with various genes that help them overcome the immunosuppressive and metabolic stress of hypoxia. Using Maestro Z, we’ve been able to compare their function in real-time against standard CAR T cells under hypoxic conditions, and the differences are quite clear—our engineered CAR Ts perform significantly better.
I’m really excited about this data and I hope to share it in a future presentation.
Charlotte Barker:
Is the vector amount and/or the starting cell number the same in non-activated versus activated T-cell protocols, or do you use more lentiviral vector for non-activated cells?
Dr. Saba Ghassemi:
Generally, we use a higher MOI (multiplicity of infection) for the non-activated T-cell protocol compared to the activated one. This is because non-activated T cells are inherently less permissive to transduction.
That said, we've been actively working on improving transduction efficiency for non-activated T-cells using several simple yet effective methods. With these improvements, we've been able to bring the MOI closer to what's used in the activated protocols, which is a significant step forward for consistency and scalability.
Charlotte Barker:
Referring to Slide 14, what happened to the ACT-free MISO-IL18 (yellow line)?
Dr. Saba Ghassemi:
Good memory—that was likely due to IL-18–related toxicity. While we didn't conduct a formal toxicity study in that specific case, based on our observations, we believe the IL-18 expression may have caused toxicity, which led to the result seen on that slide.
Charlotte Barker:
Have you seen a trade-off between initial potency and long-term persistence when comparing activated versus non-activated CAR T cells?
Dr. Saba Ghassemi:
No, we haven’t observed a trade-off. In fact, non-activated CAR T cells not only show strong initial potency but also exhibit better long-term persistence. This has been consistent across multiple experiments. The non-activated cells seem to maintain a more stem-like or memory phenotype, which likely contributes to both their durability and efficacy over time.
Charlotte Barker:
How well does the in vitro data from Maestro Z match with your mouse studies?
Dr. Saba Ghassemi:
Very well—they correlate closely. One of the key advantages of the Maestro Z system is that it provides real-time killing kinetics, which gives us insight into response speed and magnitude—something endpoint assays don’t capture as effectively. What we observe in vitro generally predicts in vivo outcomes reliably, especially in terms of which groups perform better.
Part of the potency is the persistence—and non-activated CAR Ts are very potent. You know, they are potent. They sometimes have a bit of a lag just to start, but that’s really just a matter of the CAR T cell number. But they are very important, and in some cases, like in the one-day activated versus non-activated comparison, they’re even faster than the activated ones. So, there is no trade-off.
Charlotte Barker:
Perfect, thank you! Okay, great—thank you very much, Saba, for answering those questions. That is all we've got time for today.
As today was live, April 30th, any questions we didn’t get to—we will reply to by email, so don’t worry, we will get back to you. The webinar will be available on demand starting tomorrow, so look out for an email from us with the link. We’d also like to invite you to take a short survey to provide feedback on today’s webinar. A link to the survey can be found in the Resources section on the right-hand side of the screen, and it will also be included in the on-demand email.
There’s also an attendance certificate available, which can also be found on the right-hand side of your screen.
So, all that’s left is to thank Saba once more for a great presentation, and thank you very much to the audience for joining us today. We do hope you'll join us again soon!
