Cancer immunotherapy has made incredible progress over the last decade, thanks to strategies like checkpoint inhibitors and cell therapies that tap into the power of T cells, the most effective assassins in our immune arsenal. Other immunotherapies like cancer vaccines and oncolytic viruses also work in part through their ability to boost T cell activity.
To help more patients, we need to know more about these complex cells and what makes them tick, and that’s exactly what CRI scientists and others explored at the 2021 virtual annual meeting of the American Association for Cancer Research (AACR21). In general, the fates of T cells within tumors—which impact the fates of people with cancer—are determined by two things: the cell’s internal signals (or lack thereof), and the signals from its surroundings, known as the tumor microenvironment or TME.
Regarding internal T cell mechanics, Anjana Rao, Ph.D.—a former CRI fellow who is now a CRI Scientific Advisory Council member at the La Jolla Institute for Immunology—highlighted her team’s work showing how knocking out genes for certain pathways in chimeric antigen receptor (CAR) T cells could improve their anti-cancer activity and the survival of mice. Rao, who is a professor at the La Jolla Institute’s Center for Autoimmunity and Inflammation and Center for Cancer Immunotherapy and also received CRI’s 2016 Frederick W. Alt Award, explained that this enhancement was linked to epigenetic differences in CAR T cells, meaning that different genes were accessible in order to be turned on/off. This led Hyungseok Seo, Ph.D., a CRI-Donald J. Gogel Fellow in Rao’s lab, to investigate the important elements at play. His screening experiments found that a transcription factor called BATF was particularly important. When CAR T cells were made to produce more BATF, they expanded more effectively and improved mouse survival. They also maintained an activated effector state rather than becoming exhausted, and established immune memory that enabled them to protect against future tumor re-challenges.
Andrea Schietinger, Ph.D., a CRI Lloyd J. Old STAR at Memorial Sloan Kettering Cancer Center (MSKCC), further explored the different fates of T cells. Schietinger, an associate member of the immunology program at the Sloan Kettering Institute, began by discussing how the behavior and behavioral flexibility of T cells within tumors can change over time. Ideally, when the immune system responds to acute threats, naïve T cells become effector T cells which then become the memory T cells that provide long-term protection should the threat return. But if the threat persists—like with tumors—the chronic stimulation can overwork T cells and cause them to become exhausted instead.
During the early stages of tumor development, T cells still experience exhaustion, but their identity remains plastic and reversible. This means that their anti-tumor activity could still be rescued via checkpoint immunotherapy. In contrast, many T cells in advanced, established tumors appear to be fixed in their exhausted state, and resist reprogramming even when immunotherapy is applied. This more permanent dysfunction is also associated with a variety of immunosuppressive factors in the TME, including certain immune-silencing cells, inhibitory pathways like immune checkpoints and TGF-beta, and physiological factors like elevated acidity, low oxygen, and poor nutrient levels.
Even within advanced tumors, though, there are some subsets of T cells that appear to remain responsive to immunotherapy. In this regard, Schietinger examined how the strength with which T cells bind cancer cells influences the T cells’ fates.
Unsurprisingly, T cells whose T cell receptors (TCRs) don’t bind strongly to a cancer antigen send out too weak a signal and don’t get activated, although they retain latent functionality. Interestingly, T cells whose TCRs bound to cancer cells too strongly also didn’t fare well, as the resultant signal caused them to become overstimulated and eventually too exhausted to kill cancer cells. This was also associated with persistent signaling in the TOX pathway, although Schietinger discovered that simply knocking out TOX wasn’t ideal either, as T cells lacking it completely died right after activation. Her work was also able to discern TOX’s specific impact, suggesting that while it is required for T cell persistence, it wasn’t linked directly to T cells’ killing capabilities.
Using a clever engineering approach, Schietinger developed “Goldilocks” T cells with activation signals that were just right. Importantly, this didn’t require decreasing the TCR-cancer antigen binding strength. Instead, she used T cells with high-affinity TCRs that strongly bound the cancer antigen, and then dampened the subsequent signal sent inside the T cells, so that they weren’t too taxed and were able to maintain their activated and functional state.
CAR T cell pioneer Michel Sadelain, M.D., Ph.D.—the current director and Stephen and Barbara Friedman Chair of MSKCC’s Center for Cell Engineering as well as a member of the Clinical and Scientific Advisory Committee for the CRI Anna-Maria Kellen Clinical Accelerator—also highlighted a number of potential approaches that could improve the effectiveness of T cell therapies by taking into account more of the complexity of their internal signaling mechanisms.
The recipient of CRI’s 2012 William B. Coley Award first demonstrated how using CRISPR to precisely insert the CAR-encoding genes into the proper segment of the genome (i.e., where the normal T cell receptor genes reside) could enable uniform expression of CD19-targeting CARs on T cells, leading to more potent T cells and improved survival in mice with acute lymphoblastic leukemia (ALL). This more natural incorporation of the CAR code also prevented the persistent activation signaling—and eventual exhaustion—that can occur in existing CAR T therapies. Instead, after CAR activation, the receptors were recycled and then re-expressed when needed, like the normal receptors on T cells. Another complementary strategy Sadelain spoke about here was to disrupt one of the proteins responsible for “closing” certain regions of the genome after T cell activation. This in turn kept these important genes accessible, enhancing the effectiveness of CAR T cells, including their ability to protect against future tumor re-challenges.
Like his MSKCC colleague Schietinger, Sadelain also investigated strategies to fine-tune the internal activation signals relayed within T cells after CAR binding and stimulation. Here, he highlighted that removing some of the redundant secondary signaling domains was able to better balance the effector and memory programs of CAR T cells and enhance their therapeutic activity against mouse models of both mesothelin-expressing ovarian cancer and MUC1-expressing breast cancer.
Sadelain also showcased a new type of receptor—dubbed HIT—with the potential to combine the best of both worlds as far as TCRs and CARs go. These HIT complexes combine the internal signaling mechanisms of natural TCRs with the ability to bind cancer antigens even if they aren’t presented in the context of the major histocompatibility (MHC) system, like CARs. Perhaps most importantly, these HIT receptors provided engineered T cells with greater sensitivity to cancer antigens. This could help address cell therapy hurdles in blood cancers, where cancer cells can decrease expression of the target antigen to escape CAR T cell detection, and cell therapies against solid tumors, which are often very heterogeneous and contain cancer cells that don’t uniformly express the same antigens at the same levels.
While T cells themselves are clearly at the center of the cancer immunology puzzle, the other pieces can’t be ignored. Thus, the work of other CRI scientists focused on the external factors that influence T cell behavior, and how we might be able to target them to improve patient outcomes.
First, Katelyn T. Byrne, Ph.D. of the Perelman School of Medicine at the University of Pennsylvania, presented work from a trial led by Robert H. Vonderheide, M.D., D.Phil., the director of UPenn’s Abramson Cancer Center and a member of both the CRI Scientific Advisory Council and the CRI Clinical Accelerator leadership. In this trial, sixteen patients with resectable pancreatic cancer were treated before surgery with selicrelumab, a CD40-targeting immunotherapy, either alone or in combination with chemotherapy. The CD40 pathway plays an important role in priming the antigen-presenting immune cells (APCs) that give T cells their orders as far as activating them and telling them what antigens to target. (Vonderheide also leads the CRI-funded PRINCE trial that is currently exploring a CD40 treatment, in combination with checkpoint immunotherapy and/or chemotherapy, in patients with metastatic pancreatic cancer.)
Similar to the PRINCE trial, Vonderheide’s team observed here that targeting the CD40 pathway led to increased activation of APCs in the blood, and that tumors treated with selicrelumab were enriched with respect to T cells. In addition to the increased presence of activated T cells, these CD40-treated tumors also exhibited reduced fibrosis, a tactic that tumors can use to protect themselves from the immune system by keeping T cells at bay.
Finally, Dmitriy Zamarin, M.D., Ph.D., a CRI Anna-Maria Kellen Clinical Accelerator Investigator, explored the promise of oncolytic virus immunotherapy, an approach he is currently testing as the leader of a CRI-funded trial for patients with advanced forms of ovarian, colorectal, and other cancers. Thanks to their ability to trigger inflammation and immune activity, these oncolytic viruses can be used to help heat tumors that are immunologically cold, and thus re-program the tumor environment into one that is more permissive of immune responses against cancer.
Specifically, Zamarin discussed the potential of the Newcastle Disease Virus (NDV) as an oncolytic virus platform, due to its ability to infect and destroy death-defying cancer cells that have innate immune deficiencies. When combined with checkpoint immunotherapy, NDV-based therapy led mice to reject both the tumors that were injected with the virus as well as distant lesions. In humans, another oncolytic virus therapy called T-Vec has also shown benefits against advanced melanoma when combined with checkpoint immunotherapy.
One issue Zamarin pointed out is our limited ability to determine whether these anti-cancer effects result from T cell responses against tumor antigens or viral antigens that become expressed by infected tumor cells. To that end, he stressed the need to know whether oncolytic viruses induce inflammation in a non-specific fashion, or if they stimulate tumor-specific T cell responses. Further experiments revealed that combining NDV therapy and checkpoint immunotherapy expanded the TCR clonal repertoire shared across all sites, including injected tumors, distant tumors, and the spleen. But NDV therapy only triggered infiltration of tumor-targeting T cells, but only if the virus infected tumors that expressed the specific antigen targeted by the T cells. This demonstrated the importance of the oncolytic virus actually making it into the tumor, and that oncolytic viruses can overcome heterogeneity with respect to target antigen expression within a tumor.
Zamarin also noted that tumor-targeting T cells appeared to adopt distinct functional states compared to virus-targeting T cells, which may help with future efforts to unravel their roles in the context of oncolytic virus immunotherapy strategies. Ultimately, improving the utility of this approach for patients will require a better understanding of the mechanisms that determine the balance between these different T cell populations.
For more from AACR21, be sure to read our interview with Dr. Ton N. Schumacher, of the Netherlands Cancer Institute and Leiden University Medical Center, whose career contributions earned him this year’s AACR-CRI Lloyd J. Old Award in Cancer immunology.