Day 2 of the 2017 CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference in Mainz, Germany, attended by more than 1,400 scientists from around the world, offered a focused exploration of the latest discoveries and data from scientific studies aimed at identifying biomarkers, developing new therapeutic agents, and refining adoptive cell therapies.
INTRODUCTION: THE 2017 WILLIAM B. COLEY LECTURE
Jill O’Donnell-Tormey, Ph.D., chief executive officer and director or scientific affairs at the Cancer Research Institute, opened up Day 2 of CICON17 by introducing the 2017 William B. Coley Lecture, which this year is given by the University of Chicago’s Thomas Gajewski, M.D., Ph.D., who will receive the 2017 William B. Coley Award for Distinguished Research in Tumor Immunology later this year in New York City.
Gajewski reflected on his early foray into cancer immunology. When he tried to get funding to profile gene expression in the tumor microenvironment, he was told by multiple reviewers that it was a “fishing expedition.” Fortunately, Gajewski persisted and in the decades since has made many important discoveries in tumor immunology, which earned him CRI’s highest scientific honor this year, and he continues to make key contributions to the field today.
One topic he spoke about at length was how to turn “cold,” immunologically inert tumors into “hot” ones defined by T cell infiltration, inflammation, and the expression of inhibitory immune checkpoints including PD-L1. While anti-PD-1/PD-L1 checkpoint immunotherapies typically work well for melanoma patients with these “hot” tumors, other immune checkpoints that they express, such as IDO, can hinder successful anti-tumor responses.
To that end, Gajewski has been involved in clinical trials treating melanoma patients with a combination of checkpoint immunotherapies that block both PD-1 and IDO activity. Based on a promising results from a phase I trial, patients have now been enrolled in a phase III trial testing the combination, which could pave the way for its wider adoption in the clinic. In mice, he also demonstrated the effectiveness of combining an immunotherapy that inhibits another checkpoint, LAG-3, with one that activates the co-stimulatory molecule 4-1BB (also known as CD137), which plays a role in augmenting the activity of CD4+ and CD8+ T cells as well as B cells, dendritic cells, natural killer cells, macrophages, and other components of the immune system.
Additionally, Gajewski highlighted the crucial role of dendritic cells (DCs) that express the gene Baft3 in driving cancer-targeting T cells into tumors, while also noting the ability of b-catenin signaling to prevent this important immune activity. Finally, he focused on more recent insights regarding the role that bacteria in our bodies play in enabling immune-mediated tumor elimination, and showed that the types of bacteria found in patients differed among those who responded to anti-PD-1 checkpoint immunotherapy and those who did not. He suggested that stool samples—which contain a patient’s internal bacterial populations—could potentially be used to predict which patients are most likely to benefit from checkpoint immunotherapy. One wonders if they might also be given to patients prior to treatment with immunotherapy to create an environment that is more favorable to anti-tumor immunity.
SESSION 3: TISSUE BIOMARKERS PREDICTING RESPONSE TO IMMUNOTHERAPY
Thomas Tüting, M.D., of University Hospital Magdeburg, began the third session of CICON17 by discussing the role of immune cells called neutrophils in cancer immunotherapy and exploring the hypothesis that combining T cell-based immunotherapies with those targeting neutrophil-related signaling can help provide patients with more durable responses. Using mouse models of melanoma, he showed that short-term inhibition of a signaling pathway called c-MET, typically associated with wound healing but abnormally plays a role in providing blood supply to tumors and cancer metastasis, was able to promote expansion of tumor-targeting T cells and prevent neutrophil responses that benefit tumors (by impairing their recruitment to lymph nodes), ultimately improving the effectiveness of adoptive cell therapy (ACT). Additionally, Tüting revealed in mice that neutrophils exposed to tumors that had been infiltrated by T cells alter their behavior and take on an immunosuppressive, tumor-supporting phenotype. Furthermore, short-term inhibition of c-MET made anti-PD-1 checkpoint immunotherapy more effective in several solid tumor models, which is in line with his finding that melanoma patients who don’t respond to checkpoint immunotherapy have increased numbers of neutrophils in their blood, which was also shown to correlate with elevated levels of growth factors associated with c-MET.
Next up was Drew M. Pardoll, M.D., Ph.D., of Johns Hopkins University’s Bloomberg~Kimmel Institute for Cancer Immunotherapy, who began by highlighting a technology called MANAFEST, which was developed in his lab and enables high throughput analysis of a patient’s tumor-targeting T cells in a highly sensitive and specific manner. He stressed that this approach provided his team with a new appreciation of just how large patients’ tumor-targeting T cell repertoires can be, which in turn should enhance the way doctors approach patient-specific immunotherapy strategies.
In one patient with a heavily mutated tumor (classified as having high microsatellite instability, or MSI-hi) who was treated with anti-PD-1 checkpoint immunotherapy, there was a significant expansion of the T cell populations that specifically targeted mutated proteins known as neoantigens found on tumors. Due to the increased response rates for patients with MSI-hi tumors who are treated with checkpoint immunotherapy, Pardoll recommended that MSI screening should be standard practice for all cancer patients.
In the second half of his talk, Pardoll highlighted recent results obtained through a SU2C-CRI-funded clinical trial that treated lung cancer patients with anti-PD-1 checkpoint immunotherapy prior to surgery. In this trial, neoadjuvant (pre-surgical) anti-PD-1 treatment enhanced the diversity and activity of patients’ anti-tumor T cells and frequently led to major pathologic responses that could not be observed by traditional CT scans.
Following Pardoll was Wolf H. Fridman, M.D., Ph.D., of the Cordeliers Research Centre and the Paris Descartes University, who was also the recipient of the 2010 William B. Coley Award. Fridman has been instrumental in defining how the immune composition of tumors affects patient outcomes, and at CICON17 he explained how kidney tumors (of the clear cell renal cell carcinoma variety) could be divided into three distinct groups based on the identities of the T cells that had infiltrated them. Even when it came to CD8+ killer T cells that expressed the PD-1 immune checkpoint, these cells were found to be functionally different among the three groups. Furthermore, he showed that high levels of C1q, a molecule produced by immune cells called macrophages, was associated with poorer outcomes in progression-free and overall survival in patients with kidney cancer, regardless of the tumor stage.
Next, Paola Nistico, M.D., of the Regina Elena National Cancer Institute, highlighted her recent findings from analyzing tumor-targeting T cells in melanoma patients who were vaccinated against one of two melanoma-associated antigens, either alone or after chemotherapy treatment. With respect to both the Melan-A and gp100 antigens, chemotherapy prior to vaccination led to a diverse repertoire of antigen-specific T cell receptors (TCRs) on highly tumor-reactive T cells. While the T cells targeting Melan-A expressed high levels of the PD-1 immune checkpoint, it didn’t diminish their anti-tumor activity, whereas gp100-targeting T cells expressed low levels of PD-1. After chemotherapy, both vaccines generated T cells generated that lacked expression of the co-stimulatory CD28 receptor.
On the other hand, anti-Melan-A vaccination alone led to a restricted TCR repertoire of T cells that lacked expression of PD-1 and CD28 and didn’t react to tumor cells, while anti-gp100 vaccination alone led to a diverse TCR repertoire of T cells that expressed both PD-1 and CD28. While these gp100-targeting T cells didn’t react to tumor cells initially, anti-PD-1 checkpoint immunotherapy was able to restore their functionality, demonstrating that the inhibitory function of PD-1 correlated with expression of CD28, regardless of the treatment used.
Wrapping up the session on tissue biomarkers, Niels Halama, M.D., of the German Cancer Research Center spoke about immune-related biomarkers in cases of colorectal cancer that had metastasized to the liver. In the first part of his talk, Halama reaffirmed the importance of the quantity of immune cells in the tumor environment before moving on to discuss the importance of the spatial patterns occupied by these immune cells. Here, he pointed out that the position of T cells in tumors (and relative to each other) appears to have a bearing on clinical outcomes for patients. Finally, he highlighted a platform he called MINOS (Multistage Image Analysis Workflow with Virtual Sections) that he’s been using to analyze the functionality and localization of T cells inside tumors. Respecting the author’s wishes, this blog will not go into any further detail regarding MINOS, though poster A218 in the meeting’s scientific abstracts provides some insights into this innovative imaging method.
SESSION 4: NOVEL AGENTS
Ignacio Melero, M.D., Ph.D., opened this session of CICON17 and discussed his efforts to target the co-stimulatory molecule 4-1BB (also known as CD137). Due to 4-1BB’s influence on T cell proliferation and tumor-killing activity, antibodies that activate 4-1BB have shown the ability to overcome immune tolerance of tumor antigens and are now being tested in clinical trials for several cancer types, both alone and in combination with other checkpoint immunotherapies, chemotherapy, and antibody-based targeted therapies. Additionally, 4-1BB-activating antibodies have also shown synergy with adoptive cell therapy, a combination that is currently being evaluated in a pilot clinical trial in metastatic melanoma.
The next talk was given by Özlem Türeci, M.D., of the Ci3-Cluster for Individualized Immune Intervention, who focused on immunotherapy strategies aimed at the claudin (CLDN) proteins, which are major structural components of the tight junctions between epithelial cells found along the surfaces of organs and blood vessels. One specific claudin (CLDN18.2) is normally only expressed in stomach tissue, but can become expressed in many other tissues after they become cancerous.
Dr. Türeci demonstrated that targeting CLDN18.2-expressing cancer cells with an antibody called IMAB362 was able to induce immune cell killing through two different mechanisms and could also stop proliferation at higher doses. These cancer cells could be further sensitized to IMAB362’s effects through prior treatment with chemotherapy. In addition to its effectiveness in mouse tumor models, this combination also improved median overall survival in previously untreated patients with CLDN18.2+ gastric cancer.
Türeci also highlighted the ability of a bi-specific antibody that can target (and facilitate an interaction between) T cells and CLDN6-expressing cancer cells. In mice, the treatment was able to eliminate advanced tumors and significantly improve survival by re-shaping the tumor microenvironment and increasing immune cell infiltration into tumors. Furthermore, she hopes that a platform called RiboMABs, which can be used to help patients produce their own supply of the CLDN6-targeting bi-specific antibody, may be able to overcome some of the obstacles associated with this approach and improve its clinical applicability.
Next, TCR² Therapeutics Inc.’s Patrick Baeuerle, Ph.D., presented a novel platform—dubbed TRuC™ (T Cell Receptor Fusion Constructs)—for engineering cancer-targeting T cells for adoptive cell therapy. By integrating into the natural TCR complex, Baeuerle claimed that TRuC™-T cells have the potential to engage in T cell signaling that more closely mimics that of normal T cells and may be able to provide enhanced effectiveness against solid tumors, which current CAR (chimeric antigen receptor) T cells have had challenges addressing.
As a proof of principle, CD19-targeting TRuC™-T cells were shown to exhibit powerful tumor-killing activity against a Raji subcutaneous model of lymphoma, resulting in superior survival compared to two different CD19-targeting CAR T cell constructs. Importantly, these TRuC™ T cells also expressed lower levels of cytokines, which could potentially prevent the onset of the dangerous cytokine release syndrome that has been associated with certain CAR T cell approaches. Moving forward, TRuC™ T cells targeting mesothelin (an antigen expressed in several solid tumor types) are now being evaluated in mouse mesothelioma models, where they were shown to outperform CAR T cells and even demonstrated the ability to protect mice against tumor re-challenge.
Ulrike Schindler, Ph.D., of Arcus BioSciences, spoke next and highlighted her team’s efforts to target the CD73 receptor—which can promote immunosuppression through its production of adenosine—using small molecule inhibitors. By blocking CD73’s activity, the small molecule A001421 was able to reverse the immunosuppression of CD8+ killer T cells and restore their proliferation as well as their production of interferon-gamma and granzyme B, both of which are crucial for effective anti-tumor immune responses. In addition to A001421, two other small molecule inhibitors of CD73—MEDI9447 and 11E1b—were also able to restore proliferation in CD4+ T cells. Lastly, A001421 treatment was shown to lead to a significant decrease in tumor growth in a mouse melanoma model in addition to increasing the ratio of tumor-killing CD8+ T cells to tumor-protecting regulatory T cells.
Capping off the session on novel agents was Nicholas David Huntington, Ph.D., of the Water and Eliza Hall Institute of Medical Research, who discussed strategies targeting Natural killer (NK) cells in cancer. He began by pointing out that high levels of IL-15, which is essential for the survival and activity of NK cells, typically correlate with improved overall survival in melanoma patients. In response to IL-15 activity, the most frequently expressed gene was found to be Cish, which produces the CIS protein that was subsequently shown to block the ability of NK cells (as well as CD8+ killer T cells) to respond to IL-15, ultimately impairing their anti-tumor power.
In a mouse model of metastatic triple-negative breast cancer (TNBC), removal of CIS function resulted in decreased metastasis, suggesting that CIS impedes the immunosurveillance activity of NK cells. When CIS was knocked out using CRISPR-Cas9, it increased IL-15-induced proliferation in NK cells and enhanced their ability to kill cancer cells in the lab. TGF-b was another factor that interfered with the anti-tumor activity of NK cells.
SESSION 5: ADOPTIVE CELL THERAPIES
Juno Therapeutics’ Hyam I. Levitsky, M.D., began the fifth session of CICON17 with a discussion of how to improve the benefits of genetically modified CAR T cells for cancer immunotherapy. He first addressed the problem of obtaining an accurate pharmacokinetic analysis, or understanding, of how these “living drugs” work within the body and how they affect the body and disease. He also stressed the need for tools that can quantify CAR T cell dynamics from a whole-body perspective. Optimally, this method would be minimally invasive, highly sensitive, and time-efficient, while also allowing for repeat measurements that correspond linearly to the quantity of CAR T cells in the body. Additionally, it shouldn’t impact the functionality of these tumor-targeting immune cells.
Therefore, Levitsky’s team pursued a strategy that involved engineering CAR T cells to express a shortened variant of the PSMA molecule that enabled them to track the cells using PET imaging. Importantly, CD19-targeting versions of these PSMA-expressing CAR T cells were just as capable of exerting anti-tumor activity, and a similar construct is now being developed for clinical testing. Another issue is tracking specific T cell receptors (TCRs) to determine cell fate during CAR T cell manufacturing as well as after administration in patients. To address this, they’re seeking to link TCRs to the cell-intrinsic activity of individual CAR T cells, which would enable genome-wide RNA expression profiles to be used for the assessment of individual immune receptor sequences across multiple patients.
Next, Dirk Busch, M.D., of the Technical University Munich, discussed how to select the best T cell subsets for use in adoptive cell therapy. Through single cell fate mapping in the body that used TCR signal quality as a proxy indicator, Busch’s team showed that central memory T cells (TCM) had relatively high regenerative potential even when fewer TCM were transferred into patients, a finding that was in line with previously reported data on CAR T cells.
Busch then went on to describe the first longitudinal microscopy of CAR T cells in action, meaning that the cells could be observed over an extended period of time. This analysis showed that CAR T cells could control the growth of established lymphoma tumors in the central nervous system. This coincided with uniform distribution of CAR T cells throughout the tumor. These CAR T cells also demonstrated long-term persistence (100+ days) and the ability to enter the bloodstream as well as infiltrate lymph nodes. Busch also showcased a potential safeguard strategy aimed at enabling the efficient recovery of B cells after elimination of the cancerous populations: getting rid of the B-cell targeting CAR T cells by targeting them against other CARs, a process he referred to as anti-CAR-CAR treatment.
The following speaker, Michael Jensen, M.D., of the University of Washington School of Medicine, stayed on the topic of CAR T cells, this time looking at the necessary next steps for enhancing the curative potential of CD19 CAR T cell immunotherapy for children with acute lymphoblastic leukemia (ALL). Overall, he stressed that reproducible CAR T cell products—made possible by precisely controlling their composition and differentiation—will be crucial for increasing the effectiveness of CAR T cell immunotherapies for pediatric leukemia patients. In addition to improving the reliability of CAR T cell engraftment, bi-specific CAR T cells that can target multiple antigens could also help prevent tumor escape and increase the durability of remissions in these patients.
GlaxoSmithKline’s (GSK’s) Cedrik M. Britten, M.D., spoke next about the challenges facing the optimization of engineered T cell immunotherapies, which he stressed will become broadly available in the near future. One current obstacle that stands in the way of broader patient access is the cost of these cell-based immunotherapies. Making them more affordable will require, according to Britten, innovative supply chain solutions that allow for more scalable manufacturing processes. As an example he pointed to GSK’s approach, which consists of a fully disposable viral vector platform that utilizes a stably-transfected cell line that can grow in suspension, and which he claimed has the potential to produce hundreds to thousands of doses per batch and ultimately bring down costs. The cell products themselves must also be made more automated and scalable as well as equipped with on-site processing and quality control testing.
Next, Anja Feldman, M.D., of Helmholtz-Zentrum Dresden-Rossendorf, presented work on a new type of CAR T cell platform called the UniCAR System. With respect to tumor-killing potential and cytokine production, UniCAR T cells were comparable to conventional CAR T cells, but the UniCAR system’s modularity grants added flexibility that allows dual-targeting of different antigens either simultaneously or subsequently depending on the antigenic profile of a given cancer. This can increase the efficiency of tumor cell killing as well as minimize the risk of tumor escape through antigenic evolution. The system also enables individual UniCAR T cells to be equipped with a variety of modules that provide multiple functions as well as a controllable switch that can de-activate cells and offer increased safety.
The final speaker on Day 2 of CICON 17 was Baylor College of Medicine’s Andrew G. Sikora, M.D., Ph.D. Dr. Sikora’s talk highlighted a somewhat controversial idea: that activation of TGF-b—long thought to possess only immunosuppressive capabilities—can actually serve a beneficial purpose by converting myeloid-derived suppressor cells (MDSCs) into tumor-killing immune cells. He hypothesized that TGF-b-MDSCs can be “weaponized” and deployed as an adoptive cell therapy for solid tumors, in particular for HPV+ head and neck cancer. He showed that TGF-b1 induces a unique MDSC phenotype (via SMAD2 signaling) that does not suppress T cells, but instead induces FAS-dependent tumor killing. In mice, the anti-tumor effects of TGF-b-MDSCs were amplified in combination with radiation therapy, both slowing (and nearly stopping) tumor growth as well as extending survival. Importantly, Sikora also demonstrated that TGF-b1 has similar effects on human MDSCs, therefore suggesting that this might represent a promising adoptive cell therapy approach in human patients too, although other aspects of their behavior—such as intratumoral homing properties, antigen presentation capability, and effects on the composition of the tumor-immune microenvironment—must still be investigated further.
That concludes our coverage of Day 2 of the 3rd annual International Cancer Immunotherapy Conference (CICON17). We look forward to sharing our Day 3 report tomorrow.
Photos courtesy of the Association for Cancer Immunotherapy (CIMT)/Andrea Enderlein