Sarcoma Awareness Month: Tackling Chordoma with Dr. Cassian Yee

In the early 2000s, Cancer Research Institute (CRI) Investigator Cassian Yee, MD, led an important breakthrough: under the guidance of CRI Scientific Advisory Council member Philip D. Greenberg, MD, at the Fred Hutchinson Cancer Research Center, Yee performed one of the first successful uses of adoptive cell therapy in melanoma.

Now, as a CRI-Chordoma Foundation CLIP Investigator at the University of Texas MD Anderson Cancer Center, Yee is taking the cell therapy expertise he’s honed over the last two decades and aiming at a new cancer, a rare type of sarcoma called chordoma. To do so, he’s working to better understand how T cells target tumor cells via the markers, known as antigens, that they display on their surface.

For Sarcoma Awareness Month, I spoke with Yee to learn more about how he’s tackling this challenge and what his efforts might mean for the cancer immunotherapy field at large.

Arthur Brodsky, PhD

You've been doing cell therapy since around the time CRI first funded you as an investigator in 2000. Then, you were targeting melanoma, but now through your CRI-Chordoma Foundation CLIP grant, you're going after a type of rare cancer called a sarcoma, which doesn't respond as well to current immunotherapies. In particular, you’re focusing on a type of sarcoma called chordoma.

What motivated you to go after this type of sarcoma? And why don't current checkpoint immunotherapies work well in this type of cancer?

Cassian Yee, MD

Let me just preface this by saying that the Cancer Research Institute supported our work at a time when practically very few other agencies and the National Institutes of Health (NIH) would step up to the plate. I’m really grateful to Drs. Lloyd Old and Jill O'Donnell-Tormey and everybody else who has enabled us to move forward at a time when this was a bit of a desert landscape.

Melanoma is sort of a gateway cancer for immunotherapy and adoptive cell therapy, in particular. Moving beyond adoptive cell therapy from melanoma to other solid tumors has always been a goal of our lab, and a lot of amazing postdocs and fellows and graduate students have contributed to this pursuit. One of these individuals, Seth Pollack, found that mixed round cell liposarcoma broadly expressed the NY-ESO-1 antigen. So, we began targeting that in liposarcoma as well as synovial sarcoma. Seth continues in that work, and he's a wonderful colleague now. That's very rewarding to see, to continue in the sarcoma field. I just wanted to give him a shout out because he continues working in sarcoma cell-based therapies.

When I moved to MD Anderson, they had individual clinics, even for rare tumors like sarcomas. So, Dr. Neeta Somaiah and I worked to target NY-ESO-1, but as I began to look at other hard-to-treat blood and solid cancers, I realized that there were several rare tumors that were not only resistant to checkpoint immunotherapy ineffective, but actually had no good known targets to go after with T cell therapies.

Fortunately, being in the right place at the right time, we created an antigen discovery pipeline to identify the markers that sarcoma cells expressed on their surface via the MHC system. We also ran them through a number of algorithms to get the most immunogenic ones, meaning the ones best at triggering effective immune responses. And not just for common tumors, but also rare tumors. Dr. Anthony Conley, who sees a lot of patients with chordoma, had a prior relationship with the Chordoma Foundation and we saw the opportunity to work in this area because of the CLIP grant that the Cancer Research Institute sponsored. Soon, we recognized that some of these antigen targets in common tumors were also found in rare tumors like sarcomas.

As to why chordomas are so resistant to current immunotherapies, I can't give you the entire explanation. But certainly the lack of T cells infiltrating chordomas plays a role, as does the environment that these tumors grow in, which may be rich in immunosuppressive factors and preventing checkpoint immunotherapy from having its full effect. Immunotherapy works, as long as there are T cells there. If there aren't T cells, then you really don't have a vehicle to take the brakes off the system. Our adoptive cell therapy approach supplies those T cells, but the most effective strategies will likely involve combining cell therapies with checkpoints or other immunotherapies to prevent the tumor’s evasion mechanisms.

Arthur Brodsky, PhD

So, it starts with the tumors not having many mutations, which means they don’t stimulate the immune system and T cells to respond, so then there are no T cells for the checkpoints to work on.

Your work seeks to identify what the sarcoma cells look like—what targets they express on their surface—so that you can reverse engineer tumor-targeting T cells and create an immune response from scratch.

How is this work going? And are the targets you’re going after the ones that are potentially shared between different patients and cancer types, like NY-ESO-1, or are they the more unique neoantigens that would be different for each patient?

Cassian Yee, MD

Neoantigens certainly contribute to the immunogenicity of a tumor, tumor mutational burden (TMB), and patient responses to checkpoint immunotherapy. However, equally important, or at least as important, are non-mutated tumor antigens. CRI has played a major role in promoting the discovery of one group of these antigens, known as cancer-testis (CT) antigens. These include NY-ESO-1, PRAME, and MAGE antigens. These shared antigens are non-mutated and are selectively expressed in tumor cells and also general tissues, but that makes them relatively safe to target.

There are other tumor-associated antigens, like the melanocyte-associated antigens MART1 and GP1, that are overexpressed in tumors and only minimally expressed in some tissues. We can still use T cells that recognize these antigens to treat and cure patients without serious toxicities. Then, there are tumor antigens that are associated with some functional property that enables the tumor to outgrow its normal tissue counterparts or gives it some advantage with respect to growth or metastasis.

A number of overexpressed tumor-associated targets, including CT antigens, are linked to immune resistance and tumorigenicity, even though they aren’t mutated. But we still need to show that they are present in sufficient density on the tumor cell surface to elicit a T cell response.

Arthur Brodsky, PhD

Are you're looking for primarily the MHC class 1-associated antigens targeted by CD8 “killer” T cells, or also the MHC class 2-associated antigens targeted by CD4 “helper” T cells?

Cassian Yee, MD

Good question. We've had a long history of looking at both CD8 and CD4 T cell responses, and have demonstrated that a therapy using CD4 T cells targeting NY-ESO-1 could cure a patient with metastatic melanoma. Since then, we've continued to look for methods to elicit CD4 T cell response. However, the most productive approach right now involves the class 1 antigens targeted by CD8 T cells. The class 2 antigens targeted by CD4 T cells have been a little bit more difficult to nail down in terms of their immunogenicity. But I think we have the tools and infrastructure to do that.

The important thing is that no one T cell type can go at this alone, and we’d like to improve our ability to tap into the power of helper T cells, which produce a molecule called IL-2 that aids the survival of killer T cells. As it turns out, there may be ways to circumvent them. For example, we’ve created a way to generate killer T cells that in a certain state—called early central memory—that enables them to persist and proliferate long term in patients without the need for this IL-2 produced by helper T cells. Obviously, there are many strategies, many of which were pioneered by investigators supported by CRI, to engineer T cells with long-lasting function.

Arthur Brodsky, PhD

A lot of the challenges you've been discussing don’t necessarily seem unique to sarcoma. Once you validate this method and the technology is proven, it could presumably be used for other types of cancers, too, right?

Cassian Yee, MD

Yeah, I think. It kind of goes both ways. When we first identified the epitopes and antigens, they were in common solid tumors and when we cross checked we found we could use them in chordomas and other rare tumors. That provides a proof of principle for the feasibility and the effectiveness of chordoma-reactive T cells, and since that antigen is also shared across many different solid tumor types, you actually have an entree into using the same target to go after other cancers because you've already de-risked it in chordoma. In a way, using rare tumors such as chordoma that have an unmet clinical need could be a gateway to solid tumor therapy by helping to propel clinical trials and initiate collaborations. That would be a fantastic mechanism for getting the more of these T cell-based strategies into the clinical arena.

As I mentioned, there are different modalities. First came harvesting and expanding naturally occurring tumor-targeting T cells, known as tumor-infiltrating lymphocytes (TILs), as we did with melanoma back in the early 2000s. Now, once we know the antigen we want to target, we can just take immune cells from the blood and then teach them how to target that antigen. This can be done by engineering the T cells to express a chimeric antigen receptor (CAR) or, as in our case, a customized version of the normal T cell receptor.

The advantages of this method are that you can design a T cell receptor or a CAR to target anything. For TILs, you need access to a tumor that has been infiltrated by tumor-targeting T cells. And then you have to isolate those very rare, tumor-reactive T cells. It took about fifteen years of development to be able to do so successfully.

Arthur Brodsky, PhD

And expanding that to a dose size would be its own challenge.

Cassian Yee, MD

Exactly. And then we can expand them, with uniformity, to billions of cells. That allowed us to examine both the reasons for success and failure in a more rigorous fashion, and to be more flexible and agile in how we move this beyond melanoma.

Arthur Brodsky, PhD

With newer genomic engineering technologies, we can engineer T cells with a variety of properties, including the ability to resist certain immunosuppressive signals. Can you talk a little bit about the potential of these technologies to improve the effectiveness of cell therapies, either alone and in combination with other treatments, including checkpoint immunotherapy?

Cassian Yee, MD

Synthetic biology has moved the field into areas that we couldn't have imagined a few years ago, and they're actually being put into practice. I see it as a two-way information synergy because there are properties of naturally occurring T cells that people are trying to emulate in the engineered T cells. And there's engineering that can enhance the function of the transferred antigen-specific T cells because we already have a T cell receptor that we know works well and has feedback mechanisms that allow the cells to persist. We can then augment the pathways that enhance the durability or even stealthiness of these transferred T cells, which will be important in getting these beyond a boutique-type approach to a more off-the-shelf streamlined approach.

When it comes to combinations, we’re using these uniform populations of clearly defined T cells to develop a transferable cellular biomarker. Because we’re infusing a bunch of cells that we know are tumor-reactive and can traffic to tumors anywhere, we can interrogate them in ways we can’t with naturally occurring T cells. And we can examine how they behave alone as well as in combination with other treatments, whether it's a vaccine, radiation, a checkpoint inhibitor.

You just pull the cells back out and now you can examine the population that you want to influence and get to eliminate the tumor. You can examine their ability to avoid exhaustion or proliferate in the right place at the right time, and how these dynamics ultimately impacts tumor elimination.

In medical school, I learned that patients who develop infectious mononucleosis have just a few antigens, a few populations of T cells that expand 10,000, 100,000-fold, and your spleen, gets super enlarged with them. In cancer, there's a potential for these T cells to expand to much greater levels than what we currently see, to work in even more robust fashion. We just have to figure a way to unlock that and to prevent the tumor from suppressing that mechanism.

Arthur Brodsky, PhD

With viruses, there aren’t as many antigens for the immune system to target, so it focuses its response on fewer. But tumors have so many potential targets, so the immune response can get diluted a little bit. In that way, your approach mimics the viral response, in that it’s an amplified response against fewer targets, hopefully the best ones.

Cassian Yee, MD

At least initially, yes. We’re also focused on tracking and understanding how a tumor responds to a high pressure, focused response.  Ultimately, because tumors are often heterogeneous, meaning that not all the tumor cells are identical, you probably want that immune response to spread to at least a few other antigens—a phenomenon known as antigen spread—to ensure a more comprehensive attack. But you also don’t want it to spread too much to the point it dilutes the response. No question that giving T cell populations that target one, or only a few antigens, may not be enough. But our team and others have shown that if you can induce antigen spreading, it could potentially be enough. You just need other strategies to help that along.

Arthur Brodsky, PhD

Looking back how far the field has come over the last twenty years, could you share some of your thoughts about this journey, and in particular, CRI’s role in advancing the field of immunotherapy.

Cassian Yee, MD

When CRI started funding our work, you could fill my office with all the audience members attending that talk. At that time, checkpoint immunotherapy had not been put into clinical practice. We had a few scant antigens. I look back at the work that other people have done. Dr. Thierry Boon at the Ludwig Institute. Lloyd Old, obviously, at the Cancer Research Institute. My mentor Phil Greenberg, and others. You think about how you go back to first principles to understand the basis for what we're doing now.

So, a lesson I hope that people will take to heart is that please go back to first principles and understand the T cell biology. What the mechanisms are, don't go randomly overengineering or devising new protocols unless you understand some of these first principles. Not that those studies aren't very, very important also. But those are the edges of the battle. To get to where the strategies are going to arise from really involves some additional re-reading of even past literature, and appreciation of all the basic understandings of T cell biology and the ability to be flexible enough to understand that those teachings may not always be accurate. Always questioning that.

Lastly, as I was going through the process of moving from bench to bedside—which, for cell therapy, was unquestionably one of the most rapid transitions that one can imagine—I was also training graduate students and postdocs and physician-scientists. And this is something that I learned through CRI and other funding agencies. It’s critical that we support these individuals at a very early stage, and provide them the freedom to free associate, to try things that happened purely from serendipity. I can't explain enough how important that is, especially in the current funding climate. The ability to pursue a vision that may be completely imaginary, yet grounded in science, can lead to real change.

CRI has supported our lab at every critical stage so far. Unfortunately, we had no NIH funding for some of these seminal studies. The first time combination trial with ipilumumab, the first FDA-approved checkpoint immunotherapy, was a CRI-funded trial. Even getting the drugs for us to do that combination was really through the efforts of the Cancer Research Institute. Putting us together with pharma allowed us to use ipilimumab before it was approved. It was an unimaginable barrier to overcome at the time. Now, we happen to be working in chordoma in this next step beyond melanoma to rare tumors. This might influence the next generation of researchers in that area of clinical discovery, and it was catalyzed by this CRI-Chordoma Foundation CLIP grant.