Meet Dr. Yvonne Chen, One of Our New Lloyd J. Old STAR Scientists

Earlier this year, the Cancer Research Institute (CRI) launched its ambitious Lloyd J. Old STAR Program, named in honor of the “Father of Modern Tumor Immunology,” who served as CRI’s founding scientific and medical director from 1971 to 2011.

Old’s bold vision helped CRI build the foundation upon which immunotherapy has achieved its current success and helped transform how we treat cancer. Now, to bring more effective treatments to all cancer patients, we’ll need to venture beyond what’s already known and push the boundaries of what’s currently possible with immunotherapy. This will require taking risks, and that’s exactly what these STARS—Scientists Taking Risks—will do.

With CRI support—each STAR will receive $1.25 million over the next 5 years—these promising STARs are exploring high-risk, high-reward ideas with the potential to produce transformative leaps forward that will enable the field’s next great advances and bring us ever closer to realizing our vision of A Future Immune to Cancer™.

One of these promising STARs is Yvonne Chen, PhD, of the University of California, Los Angeles (UCLA), who has made a number of important advances in the realm of T cell engineering, including strategies to overcome cancer’s ability to evade and suppress the immune system.

Dr. Yvonne Chen of UCLA. Photo by Reed Hutchinson

Currently, Dr. Chen is a co-director of the UCLA Jonsson Comprehensive Cancer Center Tumor Immunology Program and an associate professor in the department of microbiology, immunology, and molecular genetics at UCLA. Previously, Dr. Chen earned a PhD in Chemical Engineering from the California Institute of Technology, where she worked under Christina D. Smolke, PhD Subsequently, Dr. Chen was a research scientist at Seattle Children’s Research Institute and a Junior Fellow in the Harvard Society of Fellows before joining UCLA.

Recently, we spoke with Dr. Chen to learn more about her work and what she hopes to accomplish during the next five years as a CRI Lloyd J. Old STAR grantee.

Arthur N. Brodsky, PhD:           

Our T cells are some of the most powerful immune cells in our body. Recently there’s been a lot of breakthroughs with modifying these T cells in order to make them better at eliminating cancer. One type, which you work with, is called CAR T cells. What are CAR T cells and how are they helping to treat cancer?

Yvonne Chen, PhD:

T cells are immune cells that are naturally able to recognize cells that should not be in the body—for example, cells that have become infected or are otherwise foreign.

CAR T cells are made by taking T cells and genetically modifying them to express a synthetic receptor that we call a chimeric antigen receptor, or CAR, that enables them to target tumor cells via the antigens—or identifying markers—on their surfaces. Our work focuses on CAR T cells made from the patients’ own T cells, although others are exploring donor-derived CAR T cell applications. When we put these CAR T cells back into the patient, they will circulate throughout the body, and ideally find and kill those tumor cells.

This technology has been under investigation for almost three decades now, but 2017 was when the FDA first approved a CAR T cell therapy for human patients. These CARs target CD19, which is a protein found on the surface of all mature B cells, including the majority of cancerous B cells.

Now, there are two different CD19 CAR T cell therapies approved by the FDA, for subsets of patients with advanced acute lymphoblastic leukemia (ALL) and non-Hodgkin lymphoma. In addition, there are a variety of additional CARs currently being evaluated for the treatment of a whole range of cancers, from brain cancer to colorectal cancer to breast cancer and more.

Arthur N. Brodsky, PhD:

In B cell cancers, CD19 serves as a great target for these because almost all B cells express this protein. As a result, these therapies work in the majority of these patients and enable them to have clinical responses. But even if treatment is successful at first, sometimes the cancer cells can lose that CD19 protein. Then the CAR T cells can’t find the cancer cells and eliminate them. What are some of the strategies you’re exploring to address this?

Yvonne Chen, PhD:

What you just described is what we call antigen escape, meaning the tumor escapes from therapy by losing the antigens that the T cells are supposed to see. So in the case of CD19 CAR T cell therapy, many patients initially achieve complete remission, meaning the tumor is reduced to below detection levels. But eventually the cancer comes back, and when it does it oftentimes no longer expresses CD19.

Because the T cells can only recognize the tumors if they express CD19, once the tumor cells lose it, they essentially become invisible to the T cells. So one strategy my lab is pursuing to overcome this problem is to engineer the T cells that can recognize more than just CD19. If the T cells can simultaneously target two different antigens that are both present on a tumor cell, then the tumor would be less likely to escape because for it to successfully become invisible to our T cells, the tumor would now need to lose two different antigens. And it would need to lose them fast, before it gets caught by a T cell.

We have been working on a bispecific CAR that targets CD19 and CD20, which is another antigen found on B cells. In mice, we’ve shown that this bispecific CAR is just as effective as the FDA-approved CD19 CAR when the tumor expresses CD19. However, unlike the FDA-approved CAR, our bispecific CAR can prevent tumor relapse caused by spontaneous loss of CD19 expression by the tumor cells.

We recently opened a phase I clinical trial at UCLA in which patients with non-Hodgkin B cell lymphoma and chronic lymphocytic leukemia (CLL) are treated with this bispecific CAR T cell therapy.

Read Dr. Chen's article on smarter and stronger T cells

Arthur N. Brodsky, PhD:

Are you using this bispecific CAR T cell approach in any other cancers?

Yvonne Chen, PhD:

Yes, we have also begun to apply this approach in multiple myeloma, which is currently an incurable disease that affects more than 32,000 new patients in the U.S. every year. Here, we are targeting two antigens called BCMA and CS1. We have seen very promising results in animal models using this bispecific CAR, so that would be our next candidate for clinical translation.

Arthur N. Brodsky, PhD:

So far, we’ve been talking about blood or “liquid” cancers like leukemia, lymphoma, and multiple myeloma. These have been a great starting point for CAR T therapies because these cancer cells are usually fairly accessible to immune cells, either in the blood or the lymph nodes.

But in the more common solid cancers that occur in the lung, breast, brain, and other organs, there are additional hurdles for CAR T cell therapies. One of the main hurdles is the physical tumor microenvironment that can help protect cancers. Sometimes this environment prevents immune cells, including CAR T cells, from infiltrating tumors. Other times, even if CAR T cells can make it into the tumor, they can be shut down and prevented from doing their job.

How are you working to make CAR T cell therapy more effective for solid cancers?

Yvonne Chen, PhD:

This is a very important question because solid tumors account for 90 percent of all cancer-related deaths in the U.S. each year. There’s a tremendous amount of interest to figure out how exactly we might translate the successes that we’ve seen in blood cancers to solid tumors.

There are a few different reasons why solid tumors are difficult to treat with all cell-based immunotherapies, not just CAR T cells. One of the big barriers, as you alluded to, is access.

When we inject the T cells into the patient, they are automatically in the same physical space as blood cancer cells. When it comes to solid tumors, the T cells need to somehow find the tumor. They need to get out of circulation to reach that tumor, and once they get to the tumor, oftentimes they become dysfunctional in that tumor microenvironment.

In many cases this sort of dysfunction is a result of suppressive molecules that are produced by tumor cells that have evolved a variety of defense mechanisms. One example found in multiple types of solid tumors is the overproduction of the cytokine called transforming growth factor beta, or TGF-β.

TGF-β, at least in the context of solid tumor environments, has several effects. First, it shuts down T-cell proliferation so that the T cells don’t grow in that environment. Second, it directly interferes with the T cells’ ability to kill target cells. A variety of solid tumors are known to overproduce TGF-β; and normally when a T cell encounters high levels of TGF-β, they essentially become dysfunctional.

Knowing this, we developed a way to engineer T cells so that when they see TGF-β, instead of shutting down, they become more active. In solid tumors that overproduce and have a high local concentration of TGF-β, this should serve as a signal that tells our T cells, “You’re in an environment where you should expect to encounter tumor cells, so be ready to kill them.”

These TGF-β-targeting CAR T cells can now see TGF-β, and instead of becoming dysfunctional, they become ready to attack tumor cells. They proliferate and grow in response, and also produce a lot of immune-stimulating cytokines. These cytokines are proteins that the T cells can secrete into the environment to spur the T cells themselves to grow as well as encourage nearby immune cells to go after tumor cells.

Arthur N. Brodsky, PhD:

That’s a really interesting approach! It almost reminds me of jiu-jitsu in that you’re using the tumor’s strength—its ability to shut down T cells via TGF-β—against it. Something that would normally hinder the immune response is instead going to activate it and make the T cells perform better, which I think is really cool.

Yvonne Chen, PhD:

Exactly. Thanks!

Arthur N. Brodsky, PhD:

I understand that developing appropriate targets for solid tumors is also a bit more difficult compared to blood cancers. In leukemia and lymphoma, almost all of the cancer cells express a well-defined target, CD19. However, in solid tumors, there is usually much more heterogeneity in terms of the antigens that these cancer cells express, so there isn’t one antigen that can be used to target all of the tumor cells.

This is true between different patients, but also within a single individual. Sometimes different tumors within the same person—such as a primary tumor versus a metastatic lesion—can express different antigens and look different to the immune system. Even in a single tumor from a patient, there can be different populations of cancer cells that express different antigens.

Given this complexity, how are you and others working to develop better CAR T cell targets for solid tumors?

Yvonne Chen, PhD:

For a tumor cell to be targeted by a normal T cell or a CAR T cell, it needs to express an antigen that the T cell can recognize. An ideal antigen is a molecule that is expressed highly and uniformly in the tumor—meaning that every tumor cell expresses high levels of this molecule—but is not expressed on our healthy cells, at least those that we cannot live without.

Those three simple requirements—high expression, uniform expression on tumor cells, and absence from essential healthy tissues—pretty much eliminates everything that we know. CD19 is a lucky exception in that the only healthy cells that express CD19 are mature B cells. Human beings can live without B cells, at least temporarily with maintenance therapy.

In solid tumors, there are very few antigens that fit that bill very well. In addition, as you also mentioned, many solid tumors are highly heterogeneous. This means that if you just take a biopsy of a patient’s tumor, oftentimes the cells on the right side look different than the cells on the left side. So a T-cell product that only recognizes a single antigen target wouldn’t be able to recognize all of the tumor cells.

One way to get around that problem is the bispecific strategy I mentioned earlier. If you could engineer T cells to recognize more than one antigen, then you could begin to target tumors that are heterogeneous. Another strategy is to try to get your T cells to cooperate with the natural immune system, so that more diverse immune responses can be generated against tumors.

Read Dr. Chen's article on smarter and stronger T cells

Arthur N. Brodsky, PhD:

So, now I want to discuss some of the technical advantages of CAR T cell technologies, but first I think it’s important to clarify how regular T cells target cells. Normally, T cells can only recognize antigens if they’re displayed in a special way, via what we call the major histocompatibility complex, or MHC, system.

One advantage of CAR T cells is that they can target antigens like CD19 on the surfaces of cancer cells when they’re not presented via the MHC system. And now, you’re working to expand the pool of potential CAR T cell targets even further, by going after antigens that are only expressed within the interior of cancer cells. Could you tell us about those efforts?

Yvonne Chen, PhD:

As we discussed earlier, tumor-exclusive antigens are just very, very rare. It’s really hard to find antigens that are only found on the surface of tumor cells. But if we start to target antigens that are not tumor-exclusive, there will likely be off-target effects against healthy cells. To get around this problem of not having surface proteins that are unique to tumor cells, we’ve been trying to engineer T cells that can actually look inside the target cell before the T cell decides whether this is a cell that should be killed or not.

The way we do so is by engineering a synthetic version of the toxic protein that T cells produce to kill target cells. However, this protein is not automatically toxic.

We have designed the T cells to make this protein in a pro-drug form that needs to be turned on before it has an effect. The T cells make these proteins, and then when the T cells recognize a target cell through surface interaction, the T cells deliver these proteins into the target cell. So from the outside if the target cell looks like a tumor cell, the T cells would deliver these molecules into the target cell.

Once inside the target cell, instead of killing the cell right away, these engineered proteins actually look around for a second signal. If, and only if, they find a second signal that indicates a tumor cell, would this protein become activated and actually kill the target cell. So, the idea here is, we would have a two-step confirmation process. The T cell first looks on the outside and asks, “Does this look like a tumor cell?” If it does, then the protein goes in, and it does the second check on the inside, also asking, “Does this look like a tumor cell?” If it does, then the target cell gets killed.

By having this two-step verification process we can increase specificity and avoid killing essential healthy cells that we cannot afford to harm.

Arthur N. Brodsky, PhD:

It’s incredibly fascinating hearing about all the technologies and tools that are at our disposal nowadays and that you’re taking advantage of in your project.

Before we wrap up, I wanted to give a chance to leave us with your overall vision for your work. In particular, what do you think you’ll be able to accomplish over the next five years, and how do you hope that your work will impact the cancer treatment landscape?

Yvonne Chen, PhD:

First of all, I would like to say thank you to CRI. This is an incredible resource that’s enabling us to pursue some of the big questions in the field. Over the next five years, I envision a three-pronged attack to address some of these outstanding questions.

Cell-based therapy is complicated, in the sense that you have to engineer things at multiple levels: at the molecular level for the receptor and the proteins that the T cells make; at the cellular level in terms of making cells that are actually functional and can remain functional for a very long time; and also at a systems level in the sense that we need to think about how the T cells interact with the tumor once they’re inside the body.

So, for the next five years we’re thinking about strategies to address questions at each of these levels.

At the molecular and receptor levels, what sequences or what structures make a good CAR? Why do they make a good CAR? What are the mechanisms behind it that would then allow us to do truly rational design of additional novel CARs that target a variety of antigens? So far, even though the field has been working on this for almost thirty years, the design of CARs is still largely dependent on trial and error. We know certain things work, and we start making minor changes to them to try to get better and better.

At the cellular level we and many others have been interested in trying to figure out what are the most important parameters that influence how effective and how functional a T-cell product is. Some of this is based on the T-cell biology, and some of it may be adjustable through outside factors such as certain nutrients or signaling inhibitors that we can add to the T-cell culture to influence their function in the long term.

By exploring factors like metabolism, we are seeing if we can influence the way T cells are modified outside the body to make sure that they remain highly functional for a long period of time once they’re deployed into the patient.

Finally—and I alluded to this earlier—we’re very interested to see whether we can engineer T cells to communicate with the natural immune system so that we don’t have to rely entirely on the engineered T cells to eliminate the tumor, because a lot of times the tumors are highly heterogeneous and it will not be possible to engineer T cells that can recognize every single tumor cell inside a body. So we want to figure out how to engineer T cells that have anti-tumor activity but can also interact with the native immune system so that they can recruit a robust immune response against the tumor once the T cells arrive.

By combining these different aspects, my hope is that in five years we will have a much deeper understanding of how to make T cell therapies more effective. And when they do fail, we can understand exactly why they fail and how we might make them better.

Ideally, we would want to increase the consistency of these T cell products such that we can achieve high levels of clinical benefit and safety while bringing down the cost. One aspect that my lab doesn’t focus on is systematic process engineering of T-cell manufacturing. So my hope is that, in five years, the field as a whole has a much better grasp of precisely what makes for an effective T cell therapy and how we get it done in a way that provides consistent and affordable care for the patients that need them.

Read Dr. Chen's article on smarter and stronger T cells

All photos courtesy of Dr. Yvonne Chen and UCLA