Like miniature Terminators, chimeric antigen receptor (CAR) T cells seek out and destroy cancer cells autonomously. But sometimes these cancer targeting assassins — which scientists engineer to recognize specific cancer cell markers — get a little out of control.

“We've got CAR T cells expanding exponentially,” said Travis Young, a molecular biologist who leads the Biologics division of Calibr, the translational research arm of Scripps Research Institute. “If they expand too rapidly and during that expansion, release a lot of cytokines and have a major inflammatory response, then you end up with cytokine release syndrome (CRS)… or neurotoxicity.”

While clinicians can treat patients who develop CRS or immune effector cell-associated neurotoxicity syndrome (ICANS), severe cases can be fatal (1). With his background in synthetic biology and protein engineering, Young and his team developed a strategy to prevent CAR T cell runaway expansion: switchable CAR T cells.

From his start helping his college roommates with their chemistry homework to figuring out how to design a clinical trial with brand new technology, Young hopes to make CAR T cell therapy safer for patients. With six out of nine lymphoma patients already showing a complete response to his CAR T cell therapy in an ongoing phase 1 clinical trial, Young and his colleagues are just getting started with demonstrating what switchable CAR T cells can do.

How did you first become interested in science?

One year when I was a kid, my grandfather, who worked at Brookhaven National Laboratory on Long Island where I grew up, got me a chemistry set for Christmas. But I never really picked up chemistry until I got to college. I enrolled as a computer science major because I thought that was going to be where all the jobs were. In my second year, my roommates were all doing the New York Times crossword puzzle, and I was doing their organic chemistry homework. I realized then that I had a knack for the sciences. I could probably do chemistry better than I could read and write. I changed my major to biochemistry, and I really enjoyed it.

How did you end up working on CAR T cell therapy?

I earned my PhD at Scripps Research Institute with Peter Schultz where I worked on a platform called unnatural amino acid mutagenesis. It allowed us to use a 21^st amino acid in the genetic code to endow proteins with unique chemistries and functions. From there, I was a postdoctoral researcher in Christopher Walsh's lab at Harvard Medical School in natural products. I became a gene jockey there. I became very proficient at extracting genes and cloning them in different cassettes. When I started my lab at Calibr, I had a lot of experience in molecular biology, synthetic biology, and protein engineering that lent itself to developing our switchable CAR T cell platform technology.

Why create switchable CAR T cells?

CAR T cells are remarkable. Endowing a cell with a gene that allows it to seek and destroy tumors and allowing the patient cells themselves to be the drug is a very powerful approach. Because it's a completely different mechanism from traditional chemotherapies, it is effective against chemotherapy-resistant hematological cancers.

That difference leads to new problems, however. When we genetically engineer a CAR T cell and transfer it back into a patient, it is autonomous. The CAR T cell expands when it finds tumor cells so that it can find and lyse more tumor cells. That expansion of the drug in the body is the opposite of the central dogma of drug development to date, which is that when we give a subject an antibody or small molecule drug, there’s a decay over time until the drug is eliminated. That presents challenges because if we don't know what the end exposure of our cells is going to be, we can end up with CRS or ICANS.

When we looked at this problem, we wanted to come up with a simple solution to leverage the potency of these CAR T cells and their abilities to seek out and destroy tumor cells, but we wanted to restore the traditional pharmacological control that we get from dosing a typical antibody or small molecule drug. That was the inception of the switchable CAR T cell program. The very basic, fundamental idea was to regain control of the cells.

How do switchable CAR T cells work?

We think of it like a software-hardware based approach because we're giving patients these genetically engineered cells — the hardware — but the cells don't do anything without software, which is the antibody switch. 
– Travis Young, Scripps Research Institute

Instead of binding to cancer cells, the CAR T cells turn on when they bind to an antibody that we create, which we call a switch. We target the antibody switches to cancer antigens such as CD19, CD20, or CD22. We deliver the engineered CAR T cells to the patient, and then when we deliver the switch, it turns the CAR T cells on when the switch sees a tumor. When that antibody switch is naturally eliminated and no longer present in the body, then the CAR T cells turn off, just like a normal drug.

We think of it like a software-hardware based approach because we're giving patients these genetically engineered cells — the hardware — but the cells don't do anything without software, which is the antibody switch. The response is easily tuned by the amount of the antibody switch that we give the patient. Now we've got all of these different levers for how we can control these CAR T cells in the body.

Have you noticed fewer CRS and ICANS events with switchable CAR T cells?

Yes, and not only can we minimize the duration of adverse events such as CRS and ICANS, but when we stop dosing the switch, the CAR T cells can rest. One of the major challenges in typical CAR T cell therapy is preventing T cells from getting exhausted. We’ve demonstrated in our preclinical models that this rest phase when there's no switch present actually makes the CAR T cells more potent and more durable over time. We're on a path to demonstrate that in the clinic as well.

We're still learning what we can do with these switchable CAR T cells and the different ways we can modulate their responses. The thing that excites me from a basic science perspective is that this is a much more physiological way of activating a T cell. A T cell responds when there's an infection by activating, expanding, eliminating the infection, and then contracting. They then enrich a central memory CAR T cell population, which is ready to protect against subsequent infections. That's the way that this switchable platform is intended to work, except that the antibody switch now causes the activation, the rest, and the reactivation.

What was the biggest challenge you faced in setting up your clinical trial for switchable CAR T cells?

Going into it, there was a lot that we didn't know! The immune systems of the preclinical models that we tested didn’t fully recapitulate the human immune system, so it was going to be difficult to determine how to dose the CAR T cells and the antibody switch. Neither the engineered CAR T cells nor the antibody switch had been in a patient before, so we wanted to start the dose at a level that was safe. But at the same time, we didn't want to start exceedingly low because we didn't want to waste a lot of patients' time with potentially subefficacious doses. Because of the complexity and time that it takes for these people to receive therapy, we wanted to make sure that everybody had a shot at a response.

When we looked at the data from the first three subjects, two out of the three had complete responses to the therapy with the lowest doses of our antibody switch and the lowest dose of our cells. Compared to a lot of other antibody-based drugs, which are dosed at 10 milligrams per kilogram or higher, we dosed ours at just 10 micrograms per kilogram. It's 1000-fold lower than those other drugs, so it’s a very potent system.

What was your reaction to seeing those first complete responders?

When we saw that two out of the first three subjects responded, it was really gratifying. The effort from the team over the last ten years resulted in something that was, first and foremost, safe. Second, from there on out, we knew that we were going to provide every subject with the opportunity to respond to the therapy.

Are you following the same patients to monitor the durability of the switchable CAR T cell treatment?

We are. The tests are still in their early stages, but it looks promising with regards to durability. When we reported the first set of data in September 2022, we announced that the majority of the subjects had complete responses after just one or two cycles of the antibody switch. In the trial, the patients received up to six cycles, and as we administered more cycles, we saw a deepening of their response to the therapy. We saw further decreases in lesion sizes, and during that period, we didn’t see any tumor relapses. One person had a tumor come back about a year later, but the other responses were still maintained during this period. We're continuing to follow those patients, and we're optimistic about the treatment durability.

Could switchable CAR T cells also treat solid tumors?

Solid tumors are the next frontier for CAR T cells. Folks have put forward many different programs using CAR T cells to target solid tumors, but there are some challenges in solid tumors that are not shared with hematological cancers. The first is the choice of the antigen. When we engineer a CAR T cell for a B cell malignancy, the most common antigen that we choose is CD19 because it's expressed on those tumor cells. But normal healthy B cells also express CD19. So, when a subject receives CAR T cells targeting CD19, the collateral damage is healthy B cells, but people can live fairly normal and healthy lives with depleted B cells.

Because we can tune the level of activity of our CAR T cells with the switchable platform, we may find a therapeutic index where we can target HER2 on a breast cancer cell but not in healthy tissue. 
– Travis Young, Scripps Research Institute

In solid tumors, the antigens are also present on vital tissues that we can't do without. For example, in breast cancer, HER2 is a common tumor antigen to target, but it's also expressed on lung tissues and in some places in the heart. Collateral damage there is not tolerable. There are no truly unique solid tumor antigens that lack expression in vital tissues. That's why creating a CAR T cell for a solid tumor is so difficult.

Because we can tune the level of activity of our CAR T cells with the switchable platform, we may find a therapeutic index where we can target HER2 on a breast cancer cell but not in healthy tissue. We think that pharmacological control is going to be a major advantage in using CAR T cells for solid tumors. We’re working in collaboration with AbbVie applying what we've learned now with switchable CAR T cells in liquid tumors to solid tumors, and we're very excited about that partnership.

What has been the most rewarding aspect of working on this project so far?

Every time we have a tumor scan come back from a patient that still shows a complete response, that is the most rewarding thing. To completely eliminate tumors in preclinical models is exciting, but there are lots of things in preclinical models that cure cancer. Trying to recapitulate those findings in a patient is always the challenge. To see that the fundamental science behind this bore out in patients is really gratifying to me.

This interview has been edited for length and clarity.

Reference

1. Sheth, V.S. and Gauthier, J. Taming the beast: CRS and ICANS after CAR T-cell therapy for ALL. Bone Marrow Transplant  56, 552-566 (2021).