Engineering Better CAR T Cells with CRI Lloyd J. Old STAR Yvonne Chen

CRI STAR Yvonne Chen is exploring several strategies to improve the safety and effectiveness of CAR T cell immunotherapies for all cancers

Chimeric antigen receptor T cells — known as CAR T cells — are engineered cell therapies that have revolutionized care for blood cancers like leukemia, lymphoma, and multiple myeloma. However, this success has not yet extended to solid cancers, which are responsible for 90% of cases and cancer-related deaths.

As a CRI Lloyd J. Old STAR at the University of California, Los Angeles (UCLA), Yvonne Chen, PhD, seeks to change that. By investigating strategies to overcome antigen escape and the immunosuppression often present within solid tumor microenvironments, Dr. Chen is working to make CAR T cells both safer and more effective for all patients.

To learn more about her progress during her first three years as a CRI STAR, we recently spoke with Dr. Chen, who is a co-director of UCLA’s Jonsson Comprehensive Cancer Center Tumor Immunology Program, and an associate professor at UCLA in the departments of both microbiology, immunology, and molecular genetics as well as chemical and biomolecular engineering.

TRANSCRIPT BELOW

Arthur Brodsky, PhD 

Hello, I’m Dr. Arthur Brodsky, associate director of scientific content at the Cancer Research Institute. And I’m excited to be joined today by Dr. Yvonne Chen, of the University of California, Los Angeles, and the UCLA Jonsson Comprehensive Cancer Center. Welcome, Dr. Chen.

Yvonne Chen, PhD, PhD

Thank you for having me.

Arthur Brodsky, PhD 

So last time we spoke was right after you were selected as a CRI Lloyd J. Old STAR, and you laid out your vision for how you want to design better CAR T cells. Before we get into the discoveries that you’ve made, I thought it would be good to start with a brief overview of CAR T cells.

Yvonne Chen, PhD 

Sure. We work on CAR T cells–CAR stands for Chimeric Antigen Receptor, which are essentially synthetic proteins that we engineer to help redirect T cells, which are white blood cells, to recognize diseased cells that we want to get rid of, and we focus primarily on cancer cells. There are a number of CAR T cell therapies now approved by the FDA, mostly to treat hematological malignancies, leukemia, lymphoma, and multiple myeloma. And we as researchers are primarily interested in understanding what currently limits the safety or efficacy of CAR T cell therapy and finding ways to engineer better ones that can provide greater benefits to our patients. And so, the most common form of CAR T cell therapy is what we would call autologous T cell therapy. That means we take T cells out of the cancer patient, we genetically modify them to express tumor targeting CARs, we expand or grow these T cells in the laboratory, and then we give the T cells back to the same cancer patient, where these T cells can now recognize tumor cells and hopefully kill them. We’re developing newer kinds of CARs, for example, CARs that can target more than just one antigen, which would be a protein associated with the tumor. And that makes it more difficult for the tumors to escape by losing the target antigen.

Arthur Brodsky, PhD 

You alluded to the new strategies that you’re using to try to improve CAR T cell therapy for patients. And I know that one of the challenges we discussed last time was antigen escape, whereby you can design CAR T cells to target one molecule, and then the cancer cells can lose that molecule and essentially become invisible. But I understand that you’ve developed a new CAR T cell. That’s called a bispecific CAR T cell that targets two different cancer markers. Could you tell us a little about how that trial has been going?

Yvonne Chen, PhD 

Yeah, definitely. So, the bispecific CAR T cell therapy we developed is for non-Hodgkin lymphoma. And it targets two different antigens called CD19 and CD 20. These are both proteins on the surface of B cells, including the vast majority of cancers for lymphoma B cells. And the rationale for having a bispecific CAR is it makes it a lot more difficult for the tumors to escape from therapy. They sometimes can do that by losing the antigen that is targeted by the T cell. So, if the T cell can target two different antigens, the tumor would have to lose both of them in order to successfully escape, and the probability of that happening is much lower than if they only needed to lose one. So, the two targets that we have aimed our T cells at–CD 19 and CD 20–are both very highly expressed on lymphoma cells. We started a phase 1 dose escalation trial here at UCLA at the end of 2019. We have been able to treat 11 patients to date, the most recent patient was dosed about a month ago, so it was too early to assess for response. But among the first 10 patients who have reached response assessment, 9 of them responded to therapy. And 7 of them reached a complete response. That’s a very high response rate for non-Hodgkin lymphoma. And very importantly, we have seen very little toxicity that’s directly associated with CAR T cells. We have to date not seen any neurotoxicity in any of our patients. And the only cytokine release syndrome, which is another commonly seen side effect associated with CAR T cell therapy, that we have seen so far is grade one, which is the mildest grade. And so, we’re particularly excited about this because the field over the past decade or so, has come to expect that for T cell therapy to be very efficacious, we also need to accept a fairly high level of toxicity. And here we have an example that would suggest that if we engineer the therapy correctly, we may not need to accept that kind of a tradeoff. At least not in non-Hodgkin lymphoma.

Arthur Brodsky, PhD 

That’s incredible, to learn how potentially safe and effective they are. In addition to cancer being able to escape the immune system through the antigen escape that you just discussed, the CAR T cells can also become exhausted, which can be an issue for their persistence over time and cause them to become dysfunctional. And then when they become dysfunctional, the cancer can come back. Are you looking at any strategies to hopefully address this?

Yvonne Chen, PhD 

Yeah, so you pointed out a very important problem in the CAR T cell engineering field, which is sometimes we get CARs that work beautifully, such as the CD19 CAR and the bispecific CAR we’re testing in the clinic. And sometimes they don’t. And it’s not always obvious why some CARs don’t work very well. What we can observe is that some of these T cells get exhausted, as you say, and that’s actually a scientific term in T cell biology, where T cells basically become dysfunctional over time as if they have run out of steam. And it’s not always obvious, again, why the T cells are becoming exhausted. One proposed mechanism is what you mentioned earlier, tonic signaling. That refers to when these CARs, which are receptors, will signal even when they’re not supposed to be signaling. So even when there is no antigen that’s triggering the CAR, some of these CARs will just signal at baseline. And it’s believed that excessive amounts of signaling eventually drives the T cells to exhaustion. And so, as I mentioned earlier, our bispecific CAR targets CD19 and CD20. And so, one of the natural questions is why do you need both? Why not, for example, just target CD20, which is actually an antigen that has a long history of being treated in the clinic for lymphoma. And in fact, patients often retain CD20 expression even after repeated rounds of anti-CD20 antibody therapies such as rituximab. And so that would suggest to us that CD20 isn’t very vulnerable to antigen escape. And yet CD20 targeted CAR T cell therapy has not done as well as CD19, at least not in the early years of development. And so, our lab became interested in trying to understand what’s the difference between CD19 and CD20 CARs? And through that comparison, can we identify design principles that will allow us to rationally design CARs that are reliably efficacious? And so, with CRI support, we did extensive, quite systematic analysis of the CAR protein itself, trying to modify different portions of it in order to understand which part of the CAR affects the way the CAR signals and affects the CAR T cells’ efficacy. We’re certainly not the only people who’s ever asked this question. Many, many years of work by multiple groups have already elucidated certain parameters that clearly are important, for example, the binding affinity of the CAR, the size, the rigidity of the CAR protein. What we found though, was that there are very, very subtle details about the CAR architecture and the CAR sequence that can profoundly influence the way CAR T cells behave. It’s not obvious upfront that they should. To give you an example, a CAR protein is about 700 Amino acids long, give or take. Different CARs have different sizes. We found that simply inserting two amino acids–and not just any amino acids, two alanines, which are usually considered as the most benign, the most sort of inactive amino acids in these receptors–can drastically change the way they attack tumor cells. And we believe the reason that alanine insertion can change CAR T cell behavior is that it changes the shape or conformation of the CAR protein. We also found that if you change the sequence of the portion of the CAR called the single chain variable fragment, or scFv, that actually binds the antigen, can also significantly impact the CAR, but not for the obvious reason. It’s not changing the CAR by changing the binding affinity. The affinity is the same, it binds the same antigen at the same spot. And yet changing the sequence can drastically change tonic signaling behavior, as well as just the overall functionality of the CAR T cell. And so these are the kinds of investigations we’ve been able to do with CRI support and the hope is that at the end of the day, we’re going to move closer to having CAR T cell engineering be engineering, as in we understand how each part contributes to the CAR T cells’ behavior and can rationally change the different components to tune them and move farther away from trial and error, where we just have to test a lot of different designs and see which one works.

Arthur Brodsky, PhD 

It’s great to hear that you’re headed in the right direction, and as you said, more rationally designed them so you can fine tune them to do exactly what we want them to do and not do any of the things we don’t want them to do. These issues that we’ve been discussing so far are just focused on the interaction between the cancer cell and the CAR T cell. But obviously, there’s a lot of other factors at play, especially when it comes to solid tumors. You mentioned earlier that CAR T cells have worked pretty well against certain types of blood cancers. But again, solid tumors which affect all the other organs and make up the majority of tumors, don’t really work well. And a lot of that is because of the tumor microenvironment, which has signals and molecules that can shut down and suppress the immune responses, including with CAR T cells. Where is where’s the field right now, as far as the strategies that we’re exploring to tackle this issue?

Yvonne Chen, PhD 

Yeah, that’s a great question. Solid tumors account for about 90% of all cancer diagnoses and cancer deaths in the U.S. every year. And so, it is a primary focus for most CAR T cell research laboratories now trying to figure out how to make this more effective against solid tumors. There are many different things that actually contribute to the difficulty that CAR T cells have against solid tumors. For example, unlike blood cancers, which happen to be in the same physical space as T cells, which is in the bloodstream, solid tumors are not where T cells naturally are. And so, the first thing is, how do you get the T cells to find the tumor in the first place? That’s a trafficking issue. And then once the T cells find the tumor, there’s a second layer of trafficking, which is how this T cell gets through to the solid tumor, how does it T cell penetrate or infiltrate into the solid tumor? And then there’s a third problem, which is actually related to the second one, and we call that immunosuppression. That means the tumor microenvironment, the area surrounding the solid tumor, oftentimes is very hostile to immune cells. Tumor cells have learned to secrete proteins and other soluble factors that can drive T cells to dysfunction and in fact not just T cells, other kinds of immune cells as well. And so, the third challenge is how do we keep our T cells functional when they are in the tumor microenvironment. And so, my group and many others have been thinking about different strategies to overcome these various challenges. For example, one can try to make the tumor more inflamed with oncolytic viruses, etc., so that they can attract tumor cells through inflammatory signals. We can engineer the T cells to secrete chemokines and cytokines that will allow them to traffic to tumor sites. One can also engineer–this is sometimes referred to as “armoring” CAR T cells– so you have armored CARs that can secrete soluble factors to help them along in that environment. Our own group had developed, for example, a new kind of CAR that can respond to TGF beta, which stands for transforming growth factor beta, which is one of the key immunosuppressive cytokines that many different solid tumors produce. In the solid tumor microenvironment, high level of TGF beta would trigger immune dysfunction, including T cell dysfunction. And what we’ve been able to do is engineer a CAR that can not only help the T cells resist the natural signaling behavior of TGF beta, but actually turn it on its head, convert TGF beta into an activating signal for T cells. And in that way, our T cells not only do not become dysfunctional, but they also actually become more functional when the tumor microenvironment presents a lot of TGF beta. And the hope is this will not only protect our CAR T cells, our engineered CAR T cells, but actually start modifying that tumor microenvironment so that native, endogenous immune cells that we have not engineered can also benefit from our design because the environment is now changing as a result of our CAR T cell therapy.

Arthur Brodsky, PhD 

Now that you’re three years into your STAR funding, how has this CRI Lloyd J. Old STAR funding really helped you address some of these complex problems that that are at the forefront of the cancer immunotherapy and CAR T cell fields?

Yvonne Chen, PhD 

CRI funding has been tremendously helpful in our research. I think a key point is engineering cell-based immunotherapy is unfortunately very, very expensive. And so, you have to sort of decide what you are going to focus on? And what’s the most efficient way to answer that question? But sometimes science doesn’t take a straight path. The first idea you come up with doesn’t always work out. And what CRI has enabled us to do is take the risk of trying ideas that we believe are going to be good ideas, but also be able to afford the risk of it not working out, so that we’re not always trying the safest idea, something that other people have already done and we’re just doing minor twists to it. We’re able to take bigger risks, to try newer things, knowing that some of them may fail. But because we have the support of the CRI, we feel confident that we will do our best and see if we can develop a truly novel therapy that is both safe and efficacious for patients.

Arthur Brodsky, PhD 

That’s amazing to hear. And really happy to hear, as you mentioned, the high-risk work is the work that will get us to the next level and not just give us incremental improvements, but that will actually lead to revolutionary changes in care for patients and save more lives. Thank you again Dr. Chen for taking the time to speak with us today. Can’t wait to hear more about your exciting work!

Yvonne Chen, PhD 

Thank you. It’s great to be with you.