Day 3 of the third CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference, attended by more than 1,400 scientists from around the world, began with a session on combination therapies, many of which involved checkpoint immunotherapies, especially those targeting the PD-1/PD-L1 pathway, several of which have already been approved for multiple advanced cancers.
SESSION 6: COMBINATION THERAPIES
Opening the session was Genentech’s Ira Mellman, Ph.D., who first focused on combining anti-PD-L1 checkpoint immunotherapy with cobimetinib, a small molecule that interferes with the MEK pathway in cancer development. In mice, MEK inhibition by itself was shown to increase tumor expression of MHC1 (an important molecular structure that allows the immune system to recognize and target tumor cells) as well as increase survival of both CD4+ and CD8+ T cells. The combined inhibition of MEK and PD-L1 synergized and slowed tumor growth in mice more than either treatment alone. The combination is currently being evaluated in sarcoma and colorectal cancer patients, and has been found to increase CD8+ killer T cell levels.
Dr. Mellman also explored inhibition of HPK1, which can disrupt T cell receptor (TCR) signaling. When HPK1’s kinase (enzymatic) function was removed, TCR signaling and anti-tumor T cell responses improved in mice, even more so when combined with anti-PD-L1 immunotherapy. Furthermore, because poor responses and resistance to anti-PD-L1 immunotherapy are associated with a strong stromal signature characterized by high TGF-b activity, Mellman also showed the benefits of combining a TGF-b inhibitor with anti-PD-L1 immunotherapy, which increased the expression of genes related to anti-tumor T cell activity and decreased the stromal-associated genes as well as tumor growth.
The next speaker, Michael A. Curran, Ph.D., of the University of Texas MD Anderson Cancer Center, revealed the benefits of combining checkpoint immunotherapies with a treatment that seeks to reverse low oxygen levels in tumors. This low oxygen state (known as hypoxia) prevents T cells from infiltrating and attacking tumors, and has been linked to immunotherapy resistance and poor patient outcomes across multiple cancer types.
Alone, reversing tumor hypoxia enabled T cells to enter tumor regions from which they were previously excluded. Together with checkpoint immunotherapy, the combination was able to effectively treat both spontaneous and transplanted tumors in mouse models of prostate cancer, and improve survival. These effects were associated with a decrease in MDSCs (immune cells that protect tumors by suppressing immune responses), a reduction in the ability of tumors to spur proliferation of MDSCs, and an increase in the proliferation of cancer-fighting T cells. Curran is now planning a phase I clinical trial to bring this combination to patients with advanced solid cancers.
Following Curran, Memorial Sloan Kettering’s Liang Deng, M.D., Ph.D., spoke about the combination of anti-PD-1 immunotherapy with the oncolytic virus MVA (a modified, weakened strain of the vaccinia virus that is used in smallpox vaccination). Deng first showed that heat-inactivation of MVA increased production of interferon and other pro-inflammatory molecules in mouse melanoma cells.
Direct injection of heat-inactivated MVA into these mouse melanoma tumors acted as as an “on site” vaccine that cured the mice, and even elicited effective anti-tumor responses against distant tumors that weren’t injected with the virus. Short-term anti-tumor immunity required CD8+ killer T cells, Batf3-expressing dendritic cells (which are important in the overall coordination of immune responses), and signaling through the DNA-sensing pathway known as cGAS/STING. Longer-term anti-tumor memory was dependent on CD4+ T cells.
While neither anti-PD-1 nor anti-CTLA-4 checkpoint immunotherapy was effective in the melanoma-bearing mice, they each showed synergy when individually combined with the MVA injection, in both melanoma and colorectal cancer mouse models. Importantly, the MVA and anti-PD-1 combination was able to eradicate even large, established melanoma tumors in mice.
Next, Partha Sarathi Chowdhury, Ph.D., of the Kyoto University Graduate School of Medicine, explored the importance of T cell energy metabolism and mitochondrial activity. Chowdhury revealed that anti-PD-1 checkpoint blockade immunotherapy increased the mass and membrane (energy-producing) potential of the mitochondria in tumor-targeting T cells. These immunotherapy-treated tumor-targeting T cells also had higher levels of superoxide, a necessary ingredient for the production of the ROS (reactive oxygen species) that are crucial for the anti-tumor activity of T cells.
Furthermore, when Chowdhury targeted the mitochondria directly (with an ROS activator) in combination with anti-PD-1 immunotherapy, it improved the T cells’ activity against a mouse model of colorectal cancer, as did a treatment that activated a complex known as PGC1a/PPAR. Molecules known as mitochondrial “un-couplers” also synergized with anti-PD-1 immunotherapy and further increased mitochondrial mass compared to either treatment alone. Importantly, none of the mitochondrial-targeting therapies were effective against tumors in mice on their own, without checkpoint immunotherapy.
SESSION 7: OVERCOMING CHECKPOINT IMMUNOTHERAPY RESISTANCE
The next session addressed how to overcome resistance to checkpoint immunotherapy, and began with George Coukos, M.D., Ph.D., of the Ludwig Institute for Cancer Research and University of Lausanne. While checkpoint immunotherapy has thus far not achieved any significant successes in ovarian cancer, Coukos pointed out that many of these patients do have cancer-specific T cells that are activated and accumulate within tumors. To figure out how to improve immunotherapy’s effectiveness for these patients, he investigated their neo-epitope (parts of mutated proteins that can be recognized and targeted by T cells) landscape and sought to identify T cells capable of targeting these neo-epitopes.
In one study, he found that 31.5% of the ovarian cancer patients had these tumor-targeting T cells in their blood, and that they were associated with immune-reactive gene signatures. However, upon further analysis, he found that the cancer-specific T cells within the tumors were different than the ones in the blood, and in fact, that they had even higher avidity for the tumor neo-epitopes. Subsequently, he showed that vaccinating patients by exposing their dendritic cells (DCs) to tumor cells could elicit the generation of polyfunctional, neoepitope-specific T cells in addition to amplifying the pre-existing tumor-targeting T cells.
Next up was Memorial Sloan Kettering’s Alexander Rudensky, Ph.D., who discussed the role of regulatory T cells (Tregs)—the levels of which are elevated in cancer—in suppressing anti-tumor immune responses. Eliminating Tregs entirely can cause damaging autoimmunity, but he revealed that impairing Treg proliferation results in tumors that less growth and spreading to other organs.
Sergio Quezada, Ph.D., of University College London Cancer Institute, followed up Rudensky’s talk by highlighting that anti-CTLA-4 checkpoint immunotherapy has been shown to selectively deplete Tregs because Tregs have a higher density of CTLA-4 receptors compared to killer T cells. However, whether this same phenomenon occurs in humans is still a matter of debate.
Quezada developed a new model to study the role of receptors on human immune cells, called Fc receptors that interact with antibodies like anti-CTLA-4 immunotherapies. Utilizing a humanized mouse model, he showed that differences in the Fc receptor portion of the anti-CTLA-4 antibody did in fact increase elimination of tumor-infiltrating Tregs and improve anti-tumor immunity. Furthermore, data from melanoma patients appeared to confirm that differences in Fc receptors (a normal variation referred to as polymorphism) does indeed influence the response to anti-CTLA-4 immunotherapy in humans.
Quezada also spoke about targeting Tregs via anti-CD25 antibodies, and showed in mice that protection against tumor growth depended on the depletion of Tregs at the tumor site. He also developed CD25-targeting antibodies with Fc receptors specifically engineered to promote the elimination of tumor-infiltrating Tregs, and when that anti-CD25 treatment was combined with anti-PD-1 checkpoint immunotherapy, it eliminated established tumors of several different types in mice.
The next speaker was Simon Heidegger, M.D., of the Technical University Munich, who also discussed a mechanism that appears to influence the effectiveness of anti-CTLA-4 checkpoint immunotherapy. His work looked at RIG-I, a DNA/RNA-sensing receptor whose activation can cause tumor cells to die in a way (mediated by a molecule known as caspase-3) that makes them more immunogenic and supports a stronger subsequent immune response. Using a melanoma line he developed that lacks RIG-I, he showed that its activity is crucial for the anti-tumor activity of anti-CTLA-4 immunotherapy. This anti-tumor immunity also appeared to enhance the cross presentation capabilities of DCs, which subsequently led to the expansion of tumor-targeting CD8+ killer T cells that then accumulated in tumors and worked to eliminate them.
Following Heidegger, University Hospital Heidelberg’s Matthias Kloor, M.D., discussed the role of antigen presentation-related mutations in MSI-hi colorectal cancer (which is characterized by an unstable, heavily mutated genome). Due to their high numbers of mutations that make them appear “foreign” to the immune system, MSI-hi colorectal cancers often attract the attention of immune cells. However, Kloor revealed that, in 70% of these cases, the tumors have developed an escape mechanism that involves disrupting antigen presentation and limiting the immune system’s ability to target and eliminate tumor cells. A little less than half of these mutations occur in the gene that encodes the b2M protein, which interacts with MHC molecules in antigen presentation. This provides a striking example of the immunoediting phenomenon—whereby tumors are sculpted by the immune system into a form that allows them to evade detection—in action. As a result, Kloor suggested that targeting pre-cancerous lesions through neoantigen-specific vaccination, especially in those with Lynch syndrome who are predisposed to the development of MSI-hi colorectal cancer, could help to inhibit or delay the development of malignant tumors in this at-risk population.
SESSION 8: OTHER IMMUNOMODULATORS
While many of the recent advances in immunotherapy have involved those targeting the PD-1/PD-L1 and CTLA-4 immune checkpoint pathways, those treatments aren’t yet effective in the majority of patients. Thus, discovery and validation of other immune checkpoints to target in cancer patients is crucial, and that was the topic of the eighth session at CICON17.
Columbia University’s Charles G. Drake, M.D., Ph.D., tackled this idea head on, and claimed that despite the seemingly universal incorporation of anti-PD-1/PD-L1 checkpoint immunotherapies into current combination approaches, they may in fact not always be the best partner with certain cancer treatments. This assertion stemmed partly from evidence he found for adaptive Treg resistance in prostate cancer, and while he presented other substantial evidence to support his hypothesis, we respect his wishes that we not share them here (although one of his papers is set to be published soon in the journal Cancer Immunology Research).
The next speaker was Harvard Medical School’s Ana Anderson, Ph.D., who presented her findings drawn from an integrated experimental and computational approach that she developed. With this approach, she was able to identify a zinc-and-metallothioneins-containing circuit that appears to play a role in the dysfunction of CD8+ killer T cells in cancer. In CD8+ T cell populations that were previously considered uniform, her experiments—which relied upon single-cell analysis—revealed a striking degree of heterogeneity in their responses. Furthermore, her modules were able to distinguish between these different populations and their distinct functionalities.
Martin Glennie, Ph.D., of the University of Southampton, gave the next talk, which, in contrast to many of the previous talks, focused on antibodies designed to directly stimulate the immune system, rather than “release its brakes.” He highlighted a number of immunostimulatory antibodies—especially those targeting the CD40/CD40L pathway—that have shown promise both alone and in combination, and also noted (similarly to Quezada previously) the importance of the non-targeting regions of those antibodies. Interestingly, he demonstrated that Fc receptor-dependent CD40-targeting antibodies can be developed as well as those that act independently of Fc receptors. One of his most important takeaways with respect to the CD40-targeting immunotherapies is that they require clustering of CD40 and CD40L (the ligand that binds it) for it to function optimally.
Finally, Ofer Levy, Ph.D., of Compugen Ltd., capped off this session with his presentation of a novel checkpoint immunotherapy target that was identified computationally (using the platform developed by his team) and then validated biologically. The target—CGEN-15032—was found to be expressed on several different cell types, including epithelial cells (from which most cancers arise) and dendritic cells. When the gene for CGEN-15032 was “deleted” in a mouse cancer model, it inhibited tumor growth. The effect was even more enhanced when these mice were treated with anti-PD-L1 checkpoint immunotherapy.
SESSION 9: MICROBIOTA
The final session of Day 3 at CICON17 focused on the microbiota: the trillions of bacteria that reside in each and everyone one of us, most prominently in our intestines and on our skin.
Gustave Roussy Institute’s Laurence Zitvogel, M.D., Ph.D., began the session by proposing a link between gut, systemic, and anticancer immunity, and how the microbiota and its interactions with the physical barriers in the body play a crucial role in determining which way our immune scales may tip. She proposed a cause-effect relationship between anti-PD-1/PD-L1 immunotherapy resistance and an imbalance in our resident bacteria, and then went on to reveal some still unpublished data that supported her claim, and advocated strongly for the incorporation of microbiome-related biomarkers into current clinical approaches. The nature of these results emphasized a strong association between certain species of bacteria and the signaling of immune pathways that impacted anti-cancer immunity, before suggesting the potential of probiotics to normalize the interactions between the immune system and the microbiota in the context of cancer.
Following Zitvogel was the National Cancer Institute’s Giorgio Trinchieri, M.D., who pointed out that the microbiota has been shown to influence tumors in a number of diverse ways, from their genetic stability and growth to inflammation and immunity in the tumor microenvironment. The microbiota has a particularly powerful effect on colorectal cancer, perhaps unsurprising given that our intestines are where a large majority of the microbiota resides.
In one example, he revealed how bacterial digestion products can kick start an immune cascade that—via the IL-18 signaling pathway—ultimately provides protection against the development of colorectal cancer in mice. Trinchieri demonstrated that mice lacking a functional gene for the IL-18 receptor possess a microbiota that supports tumor development and could even be transmitted to normal mice through co-housing. By performing a metagenomics analysis on the microbiota found in the feces of the mice, he identified several bacterial species associated with the formation of polyps, a precursor to colorectal cancer.
Eleonora Cremonesi, Ph.D., of the University Hospital and University of Basel, wrapped up the microbiota session with a talk on the bacterial factors that influence anti-tumor T cell responses. Specifically, she presented the findings from her investigation into the chemokine networks (which control immune cell migration) responsible for driving immune cell infiltration into colorectal tumors, which is associated with improved patient outcomes. Her work demonstrated that killer T cell infiltration into tumors was associated with tumor overexpression of several chemokines, including CCL3, 4, 5, and 8 as well as CXCL9, 10, and 12. Meanwhile, CCLs 5 and 17-22, and CXCL9 and 12, were associated with the infiltration of Tregs, which can suppress the activity of the cancer-fighting T cells. Furthermore, she demonstrated that bacteria were able to trigger the tumor production of these chemokines in mice, in addition to showing that exposing tumors to gut bacteria resulted in increased T cell infiltration.
KEYNOTE PRESENTATION: THE YIN-YANG OF INNATE IMMUNITY, INFLAMMATION, AND CANCER
To top off Day 3 at CICON17, Humanitas University’s Alberto Mantovani, M.D., gave a keynote lecture that explored the double-sided coin of inflammation. In the right contexts, it helps the immune system protect the body against threats such as infections and cancer. Left unregulated, it can hasten and support the development of cancer.
Much of his talk explored immunotherapy strategies targeting macrophages, a type of immune cell that can adopt either pro-tumor and anti-tumor behavior depending on the context. He began by introducing a molecule called PTX3—which he discovered after his lab embarked on a self-professed “fishing expedition”—and describing the numerous immune-related behaviors it plays a role in. Perhaps most importantly, he highlighted its role in providing resistance against infection and its regulation of inflammation in adaptive immunity.
Probing deeper into its potential effects on cancer, he demonstrated in mice that PTX3 deficiency was associated with increased cancer-related inflammation and tumor development. These mice also had increased levels of C3 protein, which is involved in the inflammatory immune pathway known as complement. While C3 plays an important role in our innate immune system’s ability to protect us from infection, in other experiments he revealed its role in promoting tumor growth and even metastasis.
From there, Mantovani pivoted to IL-1R8, an immune signaling receptor expressed in many different tissues and cell types. While IL-1R8 was found to be upregulated in Natural killer (NK) cells during maturation, mice lacking it showed an increase in NK cell maturation, coincident with increased IL-18 signaling. The IL-1R8-deficient mice were also protected against the development of liver cancer, which depended on NK cell activity, as well as colon-to-liver and sarcoma-to-lung metastasis. This may lead to new cancer immunotherapy checkpoint therapies in the near future.
That concludes our coverage of Day 3 of the 3rd annual International Cancer Immunotherapy Conference (CICON17). We look forward to sharing our Day 4 report tomorrow.
Photos courtesy of the Association for Cancer Immunotherapy (CIMT)/Andrea Enderlein