Stanford Cancer Institute

CAR-T Cell Therapy in Solid Tumors

Handling CAR-T cells in lab

Chimeric antigen receptor-T cell (CAR-T) therapy has shown a great benefit in patients with hematologic and lymphatic cancers. However, applying the therapy to solid tumors has proven to be challenging. One of the barriers is T cell exhaustion, where the CAR-T cell becomes dysfunctional.

Many investigators are exploring solutions to this problem, and recently two publications led by Stanford Cancer Institute (SCI) members have provided new information that may help improve CAR-T therapy for solid tumor patients in the future.

Immune determinants of CAR-T cell therapy

SCI member Sneha Ramakrishna, MD, is the senior author of a recent study that used CAR-T therapy to target solid tumors expressing the molecule GD2 in patients with pediatric bone cancer and nerve cancer. Fifteen patients were enrolled. However, none of the patients’ tumors shrank. At best, some patients had stable disease. 

“In any other universe, this would be a failed trial with limited benefit for patients, but we realized that every patient we treat gives us an opportunity to learn something. This is especially true in solid tumor CAR-T cell trials, where CAR-T cells have only barely begun to provide benefit to patients. If we can see a hint of activity, we need to investigate that immune biology to understand how to improve treatments going forward.”

When they looked at the data, they saw that some patients’ CAR-T cells expanded, or multiplied, and functioned in the body. These patients had activated cytokines, indicating an inflammatory response and a sign that the body may be fighting the cancer. The team deduced that something about the CAR-T cells or the patient’s immune environment contributed to the ability of the CAR-T cells to expand in these patients.

Ramakrishna says that they were fortunate to get funding from the Cancer Immune Monitoring and Analysis Centers (CIMAC) and had access to the CIMAC’s high-quality assays, which she says were essential to yield accurate data from limited and precious patient samples. By thoughtfully interrogating patient samples throughout treatment, they were able to understand contributors to CAR-T cell expansion in these solid tumor patients.

The findings were twofold. The team identified that before being infused in the patient, the CAR-T cell products were more exhausted and less able to activate, meaning they were not as functional. Further, they found differences in the patients whose T cells expanded.   

“We found that the patients who were able to expand had healthier T cells at baseline. This aligns with findings in hematologic malignancies, but it was really the first time that we were able to characterize it in this depth in solid tumor CAR-T cell-treated patients. Seeing that the state of the baseline T cells directly correlated with the ability of the CAR-T cells to expand gave us hope to be able to optimize the CAR-T cells for these patients.”

They also found differences in patients’ myeloid cells, which have been identified as suppressive cells that stop the immune system from working properly. Evaluating these cells before CAR-T cell infusions, they found that myeloid cells expressing a cell surface protein named CXCR3 are more inflammatory and allow the immune system to function. In contrast, myeloid cells that lack CXCR3 are suppressive and stop the immune system from acting and the CAR T cells from expanding. In addition to being a marker for inflammation, the results indicate that it is also a potential driver for inflammation. 

Despite the initial differences in myeloid cells, their results showed that the myeloid cells became more immunosuppressive after CAR-T cell infusion, timed with when CAR-T cells went away in the body, even if the patient initially had pro-inflammatory myeloid cells that benefited immune system function.

“There’s fluidity across what state these cells are in, and that fluidity actually to me is really profound because it hints that maybe we can switch it and alter the immune environment to improve CAR T cell activity in patients. The beauty of this is that now that we understand this, we can think about how to make the CAR-T cells more resistant to that myeloid suppressive signature. Can we change the myeloid populations to make them more permissive for the CAR-T cells to work? How do we use this knowledge to improve the next generation of treatments that we can give our patients?” 

Ramakrishna says that the central thesis of her lab and work is the need to understand the immune biology of immunotherapy-treated patients, as it’s necessary to iterate and improve treatments, with the goal of providing patients with curative therapies.

Currently, there is an ongoing multi-site clinical trial that optimizes the design and manufacturing of the CAR-T cell product based on the study’s findings. Ramakrishna hopes that the trial will iterate on the study to improve treatment. In her lab, she plans further work to understand how the myeloid signature, its pro- or anti-inflammatory state, affects the CAR-T cells’ ability to expand and how to manipulate the myeloid or CAR-T cells to make CAR-T cells functional in the solid tumor environment. 

“Historically, people don't publish trials where there aren't beautiful responses in the patients because negative data is negative data. But, it turns out that if you really understand the negative data, it becomes very powerful and, in fact, essential to moving the field forward. This was an incredibly collaborative effort, and this group of people did an amazing job converting that data into something meaningful, important, and profound for moving the field forward. I hope that people will see our paper both for the scientific value it provides but also for the framework and approach we integrated to ask these types of questions in all immunotherapy clinical trials.”

Better nutrition for stronger CAR-T cells

SCI member Crystal Mackall, MD, is the senior author, and cancer cell therapy (CCT) senior scientist Dorota Klysz, PhD, is the first author of a January 2024 study that developed a CAR-T product showing superior anti-tumor activity in mice.

The study aimed to look at the biology behind cell-based immunotherapy and how to make it more effective for patients. 

“We weren’t setting out necessarily to make a new novel therapy. We were trying to understand the biology of how things work,” says Mackall.

Klysz focused on adenosine, a breakdown product of energy-generating adenosine triphosphate (ATP), that dampens the immune system and is produced by exhausted T cells. Adenosine exists in the tumor microenvironment and is created by tumor and immune cells. Mackall notes there are many ongoing efforts to figure out how to successfully suppress or eliminate adenosine’s activity in immunotherapy. 

Kylsz first approached the adenosine problem by knocking out three genes involved in adenosine production and signaling, but it created only a mild effect on T cells’ profile and activity.

“After that, I started to think a little bit out of the box and went back to the basic metabolic discoveries, which occurred more than 20 years ago.”

She became interested in adenosine deaminase (ADA), an enzyme that metabolizes adenosine into inosine, which is another metabolite that is not suppressive. ADA is typically present intracellularly or in lower levels on the cell surface. However, overexpressed ADA is attached to the membrane of CAR-T cells and is more readily available.

Mackall says, “Instead of inhibiting adenosine signaling or generation, Dorota said, ‘I’m just going to metabolize it by putting this enzyme there that’s going to chew it up.’ And it happens to go to inosine, but we didn’t think that was important. We just wanted to get rid of adenosine.”

Klysz explains that this manipulation gave them a strong stem-like phenotype that allows the cells to continue to renew and proliferate over time, a cell characteristic called stemness, and avoid T cell exhaustion. In mouse models, the CAR-T cells overexpressing ADA were more potent and functional when challenged with the tumor.

She says, “We observed that manipulating adenosine didn’t improve the therapy, but overexpressing ADA did. It led us to the hypothesis that it’s actually inosine that is very important and it’s driving the phenotype we’re observing.”

As inosine can substitute for glucose, which cells consume to grow, the team replaced glucose with inosine in media used to grow cells in the lab in hopes that it would create healthier CAR-T cells. This turned out to be true, as they were able to create CAR-T cells that had a similar but stronger stem-like phenotype and a greater tumor-fighting ability compared to those created when ADA was overexpressed. 

Mackall says, “Inosine is an interesting molecule because it previously had been identified as playing a role in the microbiome, and the microbiome has effects on immunotherapy. Previous investigators have done work on this, and they think, at least in some settings, some of the microbiome is producing inosine. It’s making immune cells work better. So it’s this very interesting nutrient that we are only beginning to kind of understand its immune effects, and in our hands, it was pretty potent.”

She notes that their data showed that inosine altered the CAR-T cell’s epigenome, chemical compounds that alter DNA, raising the prospect that inosine may be able to cause a more permanent alteration rather than just deliver a short-term effect. 

The study’s findings highlight the importance of metabolism in immunotherapy.

Mackall says, “There’s been very little innovation in this area. We all use the same media, no matter what kind of a cell we’re trying to grow, and we pretty much feed it the same thing. This is really highlighting the importance of the cell’s diet and provides a very cheap and easy way to alter the cell manufacturing to achieve a pretty profound epigenetic modulation.”

The study found that inosine-grown CAR-T cells grew cells at a slower rate and had more stemness, whereas glucose-fed CAR-T cells yielded higher numbers but lacked stemness.

Klysz says, “In the manufacturing field, the main discussion is always ‘how many cells can we get?’ And it’s really important because we need to get to the clinical doses, but I think one of the questions we should ask is if quality is more important than quantity.”

She explains that the next steps are to figure out which media to use with inosine supplementation and what would be the lowest dose of inosine that would induce the most favorable phenotype and potency of the cells but also wouldn’t block the cells’ proliferation. They’re also aiming to determine if inosine’s effect is transient, meaning if inosine-exposed CAR-T cells would revert back to their previous state if inosine were removed or if they’ll remain in their altered, more functional state for a long time. Another question is how long the cells need to be better for and that it might vary with the disease. For example, engineering a cell to overexpress ADA is more difficult than culturing the cell in inosine for a few days, which may be enough to get the CAR-T cell to the necessary state. 

Mackall concludes, “We started with a fundamental question and we identified something that could be useful. We were able to readily apply our discovery in process science, so we could apply it in a clinical product. The next step will be to test it in a clinical trial, and we’re in the process of determining the first clinical trial and the best setting to test it.”

February 2024
By Katie Shumake
Image: Tristan Smith from Fridays Films