CICON19 Day 2 Update: T Cell Exhaustion, the Tumor Microenvironment, and Novel Cellular Immunotherapies

The second day of CICON19 kicked off with Johns Hopkins University’s Elizabeth M. Jaffee, MD, one of the recipients of the 2019 William B. Coley Award, who gave the annual Coley lecture. Jaffee, a member of the CRI Scientific Advisory Council, began by highlighting how Coley pioneered the field of cancer immunotherapy in the 1890s. “In those days we didn’t have a lot of great science… and he made an important observation” that patients who became infected with bacteria sometimes had their tumors spontaneously disappear.

Recognizing this potential connection, Coley decided to proactively infect patients, and eventually led him to develop his eponymous Coley’s toxins. Though few appreciated the value of his insights at the time, his daughter Helen picked up his mantle and in 1953 founded the Cancer Research Institute (CRI), which has been a leader in the field of cancer immunotherapy over the last near-seven decades.

Elizabeth M. Jaffee, MD, delivers the 2019 Coley Lecture.

Due in part to CRI’s efforts, Jaffee declared that, “The field is now in a new era of precision immunotherapy.” With the help of advanced technologies, we can now analyze cancer and immune cells in incredible detail to understand what drives them—and how best to take advantage of the immune system’s power to target and eliminate tumors. Among a variety of topics and advances, Jaffee emphasized how checkpoint inhibitor immunotherapies, in particular those that target the PD-1/PD-L1 pathway, have enabled “unprecedented responses in a number of advanced cancers.” Meanwhile, newer cellular immunotherapy approaches and personalized vaccines exemplify the incredible potential of the field’s bright future.

To realize that potential, Jaffee noted that we’ll need to address several challenges. In addition to identifying biomarkers that can help doctors better determine which therapies are most likely to benefit which patients, we’ll need to develop a better understanding of how cancers become resistant to immunotherapy so that we can develop and apply strategies to overcome it.

One of the primary culprits in resistance to immunotherapy is the tumor microenvironment (TME), which is often harshly immunosuppressive and can promote “exhaustion” in cancer-killing T cells. To that end, the second day’s first session (third in the conference overall) took a deeper look at the mechanisms underlying this T cell exhaustion.

Session 3: T Cell Exhaustion: Resistance Mechanisms

W. Nicholas Haining, B.Ch., B.M.—formerly of the Dana-Farber Cancer Institute and now at Merck—discussed his search for new pathways involved in T cell exhaustion. Using the Chimeric Immune Editing (CHIME) technology his team developed, Haining identified PTPN2 as an important factor. When the gene for PTPN2—which can dampen the activity of the important interferon signaling pathway—was “knocked out,” it improved the effectiveness of adoptive cell immunotherapy. Furthermore, this deletion led to an increase in terminally differentiated T cells with enhanced cancer-killing activity, without affecting the memory-like progenitor T cell population that works to sustain the pool of terminally differentiated effector T cells. Even though these T cells appear more exhausted, their greater abundance was associated with therapeutic benefit.

W. Nicholas Haining, B.Ch., B.M., discusses Chimeric Immune Editing (CHIME) technology.

Dietmar Zehn, MD, PhD, of the Technical University of Munich, spoke next and focused mostly on TOX, a lineage-defining factor that acts predominantly in the memory-like progenitor T cell population and is critical to the induction and survival of exhausted effector cells during chronic infection and perhaps cancer, too. A lack of TOX signaling, Zehn observed, led to a decrease in the functioning of terminally differentiated effector T cells as well as a reduction in the size and fitness of the progenitor T cell population. Interestingly, Zehn also found that these effector T cells, but not memory-like T cells, depend on “helper” T cells for their activity—restoring these “helper” T cells led to the formation of new, terminally differentiated T cells with killing activity.

Dietmar Zehn, MD, PhD, discusses TOX, a lineage-defining factor.

Like Haining’s results with PTPN2, Zehn’s results with TOX reveal the seeming paradox surrounding exhausted, terminally differentiated T cells: while they are more prone to shutting down prematurely, some of them also appear to be in a state that, at least temporarily, enhances their ability to attack and eliminate infected or cancerous cells.

The following speaker—Ludovic Martinet, PhD, a CRI CLIP Investigator at the Institut national de la santé et de la recherche médicale (INSERM, France)—highlighted how the CD226 activating receptor influences T cell activity. This CD226 receptor plays a crucial role in cancer immunosurveillance and is initially expressed during the induction of effector T cells after a target—known as an antigen—is encountered. However, some T cells lose CD226 expression over time, especially in the context of cancer, resulting in less responsive T cell receptors (TCRs) that T cells use to recognize and engage the cells they’re designed to attack. Consequently, this lack of CD226 also limited the effectiveness of checkpoint immunotherapy, whereas forced expression of CD226 was able to restore the cancer-killing activity of T cells.

Ludovic Martinet, PhD, highlights how the CD226 activating receptor influences T cell activity.

Next, David Brooks, PhD, of the Princess Margaret Cancer Centre (Canada), also discussed T cells in the context of chronic infection, during which the local immune system gradually loses its ability to be stimulated by the presence of target antigens, even if they remain expressed at high levels. Brooks found that naïve T cells strongly primed in the midst of chronic infection rapidly gave rise to memory-like T cells, which in turn gave rise to effector T cells that were less exhausted. These T cells were also in a better position to respond to PD-L1 checkpoint immunotherapy, suggesting that the treatment’s effectiveness depended more upon increasing the numbers of these “already cycling” T cells as opposed to restoring the activity of previously dysfunctional T cells. Like Zehn, Brooks also showed that the restoration of “helper” T cells was an important factor in this priming that could overcome many of the characteristics of T cell dysfunction seen in the context of chronic infection.

David Brooks, PhD, discusses T cells in the context of chronic infection.

Regeneron’s Dimitris Skokos, PhD, rounded out the session with a talk on novel bispecific antibody approaches—engineered antibody-based constructs that are able to bind two targets simultaneously. First, Skokos showcased an approach utilizing two different bispecific antibodies. One binds both the T cell receptor (via CD3) and cancer cells (via MUC16, a protein overexpressed in several cancers). This treatment—which brings cancer cells into close contact with T cells—has been effective in preclinical models and is currently being explored in a Phase I trial for ovarian cancer. The other bispecific antibody also binds MUC16 in addition to the CD28 co-stimulatory receptor on T cells. Importantly, these two bispecific antibodies target different regions of the MUC16 protein, so they don’t compete with each other for binding spots. By providing the two signals—CD3 and CD28—necessary for T cell activation, the use of this CD28xMUC16 antibody enhanced the tumor-killing activity of CD3xMUC16 antibody even when the latter was administered at a suboptimal dose, suggesting potential synergy. Lastly, Skokos finished his talk by highlighting a promising bispecific antibody that targets both CD28 and prostate-specific membrane antigen, or PSMA, which is often overexpressed in prostate cancer. In combination with PD-1 immunotherapy, this treatment activated T cells, induced the expression of pro-inflammatory molecules, and promoted tumor-killing in mice without systemic toxicity.

Dimitris Skokos, PhD, discusses novel bispecific antibody approaches.

Session 4: Immunotherapies, Non-Cell-Based

Ira Mellman, PhD, a member of the CRI Scientific Advisory Council who leads the cancer immunology program at Genentech, posed a somewhat provocative question as he opened the fourth session of CICON19: Are “exhausted” T cells really “dysfunctional”—and is this reversed by PD-1/PD-L1 checkpoint blockade? After presenting interesting points that implied we may not fully understand how these immunotherapies work, Mellman demonstrated that the PD-1 pathway may inhibit T cell activity by inactivating the CD28 co-stimulatory receptor, which is bound by the protein B7.1, also known as CD80. Here, he showed that B7.1 signaling limits PD-1 signaling in T cells by sequestering the PD-L1 ligand that binds PD-1, and that inhibiting this interaction facilitated tumor growth. Mellman also revealed that within tumors, dendritic cells—immune cells that coordinate overall immune responses by stimulating T cells, but can also suppress T cell activity in some circumstances—are the primary source of this inhibitory PD-L1. Knocking out PD-L1 expression in these immune cells had a profound effect on the priming of T cells. This led Mellman to suggest, like Brooks, that PD-1/PD-L1 immunotherapies may act to enhance T cell expansion rather than to reverse exhaustion. This would mean that patients who respond to PD-1/PD-L1 immunotherapies likely have ongoing immune responses to cancer that are being held in check by the tumor and surrounding microenvironment. Consequently, we may need to reassess our treatment strategies and instead look for additional pathways that mediate immunosuppression in the tumor and that could serve as promising therapeutic targets.

Ira Mellman, PhD, discusses whether exhausted T cells are really dysfunctional.

Katharina Reinhard, PhD, spoke next regarding work she has done with BioNTech’s Ugur Sahin, MD Reinhard revealed their promising results with an RNA vaccine, incorporated into a liposomal complex, that is designed to drive the expansion of chimeric antigen receptor (CAR) T cells after they’ve been administered into the body. (The details of this work are being withheld in accordance with their privacy request.)

Katharina Reinhard, PhD, reveals promising results with an RNA vaccine.

The following speaker, Sumit Subudhi, MD, PhD, of the University of Texas MD Anderson Cancer Center and a member of CRI’s Anna-Maria Kellen Clinical Accelerator, covered the current state of checkpoint immunotherapy in prostate cancer. Unfortunately, immunotherapy is rarely effective in prostate cancer, which usually contains relatively few mutations and therefore has what is referred to as a low tumor mutational burden, or TMB. Responses are especially uncommon once the cancer has metastasized to the bone, where the tumor microenvironment can prevent T cells from entering. While CTLA-4 checkpoint immunotherapy can increase T cell infiltration into primary prostate tumors, it is often ineffective due to adaptive resistance resulting from immunosuppression. Fortunately, the combination of CTLA-4 and PD-1 immunotherapies was able to help overcome this immunosuppression in some patients, although there remains a need to explore alternative dosing and scheduling strategies to reduce the toxicity of this combination approach. Additionally, combining CTLA-4 immunotherapy with an inhibitor of the transforming growth factor-beta (TGF-b) pathway showed promise in preclinical models of prostate cancer with bone metastases.

Sumit Subudhi, MD, PhD, discusses checkpoint immunotherapy for prostate cancer.

Next, Nikesh Kotecha, PhD, of the Parker Institute for Cancer Immunotherapy (PICI), discussed his organization’s approach to deep immune profiling and informatics-driven deep analysis projects that seek to extract further value from existing data. The former focus is exemplified by the PRINCE pancreatic cancer trial that is being supported by the Cancer Research Institute in collaboration with PICI. Preliminary data from this trial, reported earlier this year at the annual AACR meeting in Atlanta, showed that the first-line combination of chemotherapy and two immunotherapies (targeting the PD-1 and CD40 pathways) led to impressive response rates in patients with metastatic pancreatic cancer. Here, deep immune profiling revealed several novel insights, including that responses were associated with the activation of B cells during treatment. As for deep analysis, Nikesh described challenges stemming from physical and conceptual fragmentation of different data sets from varying sources that must be incorporated. To address this, Kotecha’s team built a data platform called CANDEL that serves as a knowledge graph for data interpretation. Among other advantages, CANDEL is re-usable across projects due to its consistent coding interface as well as a historical record of all data ever entered, which ensures reproducibility and provides the ability to see how data and analytical outputs have evolved over time.

Nikesh Kotecha, PhD, outlines PICI's immune profiling and informatics-driven analysis projects.

As the last speaker of the session, Richard G. Vile, PhD, of the Mayo Clinic, highlighted a particularly clever approach that centered around the APOBEC3 family of proteins, which are associated with tumor formation and resistance to immunotherapy. Vile found that the APOBEC proteins help tumors escape the immune system by inducing rapid mutations in cancer cells that render them impervious to immune attack. Traditionally, Vile noted, strategies to slow mutation rates have been preferred, but in the Immunotherapy Age, driving rapid mutation in cancer cells could be helpful as it may generate more abnormal proteins for the immune system to target for elimination. This could make tumors more susceptible to checkpoint immunotherapy, as has been seen in cancers with typically high levels of mutation such as ultraviolet (UV) radiation-induced melanoma or cigarette carcinogen-induced lung cancer.

Richard G. Vile, PhD, discusses the APOBEC3 family of proteins.

Vile recognized that this also presents another, more interesting opportunity. Using a vesicular stomatitis virus (VSV)-based oncolytic virus therapy that was designed to induce APOBEC expression in cancer cells, Vile found that tumor escape was typically associated with the emergence of a mutation in the gene that encodes for the CSDE1 protein. As the vast majority of tumor cells that escaped treatment had this mutation, Vile proposed a fascinating “trap and ambush” strategy in which an initial treatment (such as this APOBEC-engineered oncolytic virus) would be used to drive predictable tumor mutations (such as mutated CSDE1) that could then be precisely targeted through neoantigen-specific vaccines or cellular immunotherapies.

Session 5: Immunotherapies, Cell-Based

Turning to adoptive cell immunotherapies, Carl G. Figdor, PhD, of Radboud University in The Netherlands discussed three strategies that might help overcome the limitations of current dendritic cell-based immunotherapy approaches. Currently, these therapies are hindered by inconsistent quality in starting material and the fact that their production is labor- and resource-intensive because they are customized individually for each patient. First, Figdor discussed the development of nanoparticles that are designed to activate immune responses in the body. These nanoparticles can be loaded with immune-stimulating factors—that target the CD3, CD28, and IL-2 pathways, for example—and their stimulatory abilities are influenced by the length of the polymer used, the density of the factors incorporated, and the stiffness of the polymer backbone. Second, Figdor spoke about building synthetic dendritic cells made of polymer filaments capable of directly inducing T cell activation. In addition to their non-toxicity, these synthetic DCs were able to migrate to lymph nodes and activate T cells there. Third, Figdor touted the most ambitious of the three approaches: synthetic immune niches designed to be injected into the body, where they can essentially mimic lymph nodes and tertiary lymphoid structures and provide local immune activation. In addition to their biocompatibility, they were also porous enough to allow for immune cell infiltration and could be readily modified to incorporate a number of immune-stimulating molecules.

Carl G. Figdor, PhD, discusses overcoming the limitations of current dendritic cell-based immunotherapy approaches.

The next speaker was Laurie Menger, PhD, of the Curie Institute (France), who discussed how she has successfully used genome-wide screens to identify novel immune checkpoints for potential therapeutic targeting. The most prominent target was SOCS1, which Menger determined to be a checkpoint found on “helper” memory T cells. She found that SOCS1 inhibits proliferation of these immune cells and is involved in a negative feedback loop that extinguishes pro-immune signaling and results in an unresponsive state. When Menger inactivated SOCS1 in these “helper” memory T cells, it restored their proliferation during ongoing immune responses, enhanced their production of a range of immune-stimulating factors (a trait known as polyfunctionality), and improved their ability to infiltrate and eradicate tumors in mice.

Laurie Menger, PhD, discusses using genome-wide screens to identify novel immune checkpoints.

Following Menger, Helen E. Heslop, MD, D.Sc., of the Baylor College of Medicine, focused on natural T cell receptor (TCR)- and synthetic CAR-based cellular immunotherapies. First, Heslop covered advances in Epstein-Barr virus (EBV)-associated cancers such as lymphoma. Her team generated T cells designed to target EBV-specific proteins expressed by cancer cells, and found that they expanded in the body after infusion and also led to antigen spreading, meaning that the immune response expanded to go after additional tumor-associated antigens (TAAs) that were not targeted by the T cells initially administered. Thus far, these TAA-targeting T cells have shown clinical benefit in the blood cancers lymphoma, leukemia, and myeloma, including when administered after surgery to prevent recurrence. With respect to CAR T cells, Heslop pointed to impressive response rates, including some complete responses, in lymphoma patients who were treated with CD30-targeting CAR T cells. Moving forward, she stressed the importance of developing “off-the-shelf” CAR T cells generated from donors, which could help lower the immense costs of this approach. Given their donor-derived origins, there is a risk these T cells would see the patient as “foreign” and attack the patient’s normal tissues, or the patient’s immune system would mount an attack against the donated T cells, and these issues must still be addressed. Two potential solutions would be to remove the normal T cell receptors from these CAR T cells, or, alternatively, to equip virus-targeting T cells with CARs to ensure that they aren’t equipped to attack a patient’s healthy cells.

Helen E. Heslop, MD, D.Sc., discusses cellular immunotherapies.

Kole Roybal, PhD, of the University of California, San Francisco, spoke next about engineering next-generation T cell therapies. Specifically, Roybal discussed the use of synNotch receptors that can sense environmental cues and then initiate specific—and customizable—cellular programs in response to those environmental cues. Per Roybal's request, the details of his promising research have been omitted from this report.

Kole Roybal, PhD, discusses engineering next-generation T cell therapies

Sessions 6: New Targets and Concepts

Ellen Puré, PhD, an associate director of the CRI Scientific Advisory Council at the University of Pennsylvania, chaired and opened the final session of the second day at CICON19. Puré focused on the “stromagenic switch” that can activate fibroblast cells and cause them to dynamically remodel the tumor microenvironment, promote the growth of tumor cells and tumor blood vessels, and suppress immune responses against the tumor. These pro-tumor fibroblasts can be generally characterized by their expression of a molecule called FAP and are especially prominent in pancreatic cancer, where they are associated with a physical, fibrotic barrier. This stromagenic switch, according to Puré, exposes therapeutic vulnerabilities in tumors and provides opportunities for the development of therapies to prevent the seeding and/or growth of metastatic cancer cells. To that end, Puré developed FAP-targeting CAR T cells that were able in mouse models to break down this barrier and disrupt the growth of tumors and their blood vessels as well as enhance the infiltration and activity of T cells within tumors. These FAP-targeting CAR T cells also improved the effectiveness of vaccination. Additionally, Puré showed that the calcineurin pathway plays an important role in preventing the development of these pro-cancer fibroblasts. When it was deleted, it promoted fibroblast migration and remodeling of the tumor matrix that altered the migration of pancreatic cells, ultimately leading to an increase in the frequency and size of metastatic lesions in the lungs.

Ellen Puré, PhD, discusses the “stromagenic switch” that can activate fibroblast cells.

Thea D. Tlsty, PhD, of the University of California, San Francisco, continued with this theme of how the stroma—which includes fibroblasts—can promote the growth and spread of cancer cells while inhibiting immune responses against them. Specifically, Tlsty focused on the role of the CD36 receptor, whose absence is associated with the flipping of the stromagenic switch referenced by Puré. Tlsty showed that fibroblasts with low levels of CD36 aided the formation of a metastasis-promoting niche in tumors, leading to an increase in the number of circulating tumor cells and metastatic lesions. Furthermore, when macrophage immune cells lack CD36, it limits their ability to phagocytose—or “eat”—cancer cells and other damaged or diseased cells. This loss of CD36 can also be induced by exposing macrophages to cancer cell-conditioned media, which converted them from cancer-fighting cells into immune-suppressing cells. Importantly, Tlsty emphasized that this loss of CD36 and flipping of the stromagenic switch is not a reactive event driven by tumors; rather it can be stimulated by several types of tissue stress in the absence of cancer. Macrophages without CD36 are also unable to clear up “debris” in the local environment, which leads to the chronic inflammation that likely aids the early development of tumors.

Thea D. Tlsty, PhD, discusses how stroma can promote the growth and spread of cancer cells.

Cancer cells can also directly protect themselves from being attacked, including attack from macrophages, according to Amira Barkal, an MD, PhD, student in the laboratory of Irving Weissman, MD, at Stanford University. Three ways they can do this—via the CD47-SIRPa, LILRB1-MHC-I, and PD-1/PD-L1 pathways—are already known, and during her talk Barkal revealed a fourth that her team recently identified. Through the expression of CD24, cancer cells can engage the Siglec-10 receptor on macrophages to prevent themselves from being eaten. This CD24 receptor is often overexpressed in several types of cancer, and is associated with poor outcomes in patients with ovarian cancer and breast cancer. When Barkal blocked this CD24-Siglec-10 interaction, it greatly enhanced macrophages’ phagocytic abilities and enabled clearance of ovarian and triple-negative breast cancer cells, promoting superior tumor control in mice with CD24-expressing tumors.

Amira Barkal reveals the role of the CD24 pathway.

The final speaker on the second day of CICON19 was Susan Kaech, PhD, a CRI CLIP Investigator at the Salk Institute. Kaech’s presentation dealt with her work on metabolic checkpoints that might serve as potential targets to enhance immunotherapy. Specifically, Kaech sought to determine where there were any alterations in nutrient availability that affect the functioning of T cells in tumors. Kaech observed that T cells within tumors often overexpressed the CD36 receptor—the same one discussed by Tlsty. But whereas Tlsty showed that CD36 makes macrophages better cancer-fighters, Kaech showed that CD36 appears to worsen T cell-mediated cancer killing. Specifically, CD36 facilitates increased uptake of already-oxidized fatty acids that can induce exhaustion in T cells and enable increased tumor growth. On the contrary, when CD36 was knocked out, it led to T cells with superior cancer-fighting capabilities. Interestingly, increased activity in several metabolic pathways—including those related to fatty acid oxidation—correlated with patient responses to PD-1 immunotherapy. Kaech provided additional evidence for this link by showing that disabling important metabolic enzymes resulted in less functional T cells and greater tumor growth.

Susan Kaech, PhD, discusses metabolic checkpoints.

That’s it for day two of CICON19. Check back later for our recap of the CICON19 highlights from day three!

Read CICON Blog Posts

Our complete CICON19 coverage: 

• CICON19 Day 1 Update: Cancer Prevention and Combination Immunotherapy Strategies
• CICON19 Day 2 Update: T Cell Exhaustion, the Tumor Microenvironment, and Novel Cellular Immunotherapies
• CICON19 Day 3 Update: Vaccines and Tumor Antigens, Technology and Informatics, and the Tumor Microenvironment
• CICON19 Day 4 Update: Metabolic Crosstalk and the Microbiome

All photos by Arthur N. Brodsky, PhD, for the Cancer Research Institute