TPS779 Background: PDA has a low tumor mutational burden and an immunosuppressive microenvironment. Targeting overexpressed self-antigens with adoptive, genetically engineered, high affinity T cells may overcome immunosuppression (Stromnes I, Immunol Rev 2014). We previously engineered murine CD8 + T cells to express a high affinity MSLN-specific TCR in the Kras LSL-G12D/+ ; Trp53 LSL-R172H/+ ; p48 Cre/+ ( KPC) mouse model, with highest efficacy observed following serial infusions (Stromnes I, Cancer Cell 2015). We developed an autologous TCR-T cell therapy targeting MSLN, which is overexpressed by 80% of PDAs. In this first-in-human phase I trial, we seek to determine the safety, preliminary efficacy, as well as the persistence, activation, localization, and functional capacity of FH-TCR T MSLN in chemotherapy refractory, MSLN + mPDA. Methods: Eligible pts have ECOG PS 0-1, ≥ 1 prior therapy for mPDA, life expectancy ≥ 12 wks, able/willing to undergo biopsies (at baseline, and after 3 and 6 wks of treatment), MSLN 2+ expression in ≥ 30% tumor cells by IHC, HLA-A*02:01, no HLA-B*13:02. Patients undergo leukapheresis, and approximately 3-4 weeks later, receive the first TCR-T MSLN cell infusion. Three TCR-T MSLN cell infusions given q21 days (d) are planned to be administered to each patient. Patients are enrolled in 4 cohorts by dose level (cohort 1: 1 x 10 9 ; cohort 2: 3.3 x 10 9 ; cohorts 3 and 4: 10 x 10 9 T cells), with a 3+3 design. Dose limiting toxicities (DLTs) are assessed during 21d after each TCR-T MSLN cells infusion. Lymphodepleting chemotherapy with fludarabine/cyclophosphamide is administered prior to the third TCR-T MSLN infusion (cohorts 1-3) or prior to the first TCR-T MSLN infusion (cohort 4). Primary endpoint is safety and DLTs. Secondary endpoints are ORR, PFS, OS. Exploratory endpoints are translational tumor and blood TCR-T MSLN cells biomarkers. Safety is assessed by CTCAE v5.0, with a goal of grade ≥ 3 TCR-T MSLN cells unexpected toxicity rate <35%; stopping rate if toxicity is ≥20% is 0.06. Response is assessed by RECIST 1.1. Enrollment of 15 patients allows >80% power to observe a statistically significant (one-sided alpha level 0.05), meaningful efficacy signal of ORR 20%. Secondary and exploratory endpoints will be descriptive and hypothesis generating. The study was activated in December 2021 and is open to accrual; 4 patients have been enrolled as of 24 Sept 2022. Clinical trial information: NCT04809766 .

Abstract While neoantigen-directed therapies have primarily focused on tumor-specific somatic mutations, emerging evidence suggests that some malignancies aberrantly translate regions of the genome outside of annotated open reading frames (ORFs), leading to HLA-I presentation of cryptic peptides. Non-canonical HLA-I bound peptides (ncHLAp) can arise from aberrant translation of genomic elements like 5’ and 3’ untranslated regions (UTRs), long noncoding RNAs (lncRNAs), retained introns, and translation in alternative reading frames. Pancreatic cancer has a low-to-intermediate mutational burden; hence, efforts to broaden the landscape targetable antigens in this disease are greatly needed. Here, we leveraged twelve pancreatic cancer (PDAC) patient-derived organoids (PDOs) to purify and enrich the malignant compartment from low tumor cellularity tumor specimens. PDOs were subjected to extensive genomic (whole genome sequencing) and transcriptomic (RNA-sequencing) profiling to enable mutation calling and HLA typing. We then employed a personalized proteogenomic platform coupled with high-depth immunopeptidomics and empirically identified >90,000 unique PDAC HLA-I-bound peptides (HLAp). We detected HLAp arising from somatic mutations (both missense and frameshift) in a subset of PDAC patients; however, we did not detect any shared mutation-derived neoepitopes in our patient cohort. Moreover, we empirically identified >1,700 ncHLAp, primarily arising from translation of novel unannotated open reading frames (nuORFs), and a substantial proportion of ncHLAp were shared amongst PDAC patients with the appropriate HLA haplotype. We developed a highly stringent pipeline using immunopeptidomics and ribosome-sequencing to investigate translation of nuORFs across a range of healthy tissues, including healthy thymus, and found that >500 ncHLAp exhibit highly-specific cancer-restricted translation. We next investigated the immunogenicity of PDAC-restricted ncHLAp and mutation-derived HLAp using an ex vivo T cell priming and expansion platform. Here we demonstrate that a subset of both mutation-derived and PDAC-restricted ncHLAp harbor robust immunogenicity. ncHLAp-specific T cell receptors (TCRs) were identified from reactive cytotoxic T lymphocytes (CTLs) via single-cell TCR-sequencing, and antigen specificity was confirmed after TCR reconstruction and cloning. Finally, using a newly developed organoid: T cell co-culture platform, we demonstrate that ncHLAp-reactive TCR-redirected T cells exhibit robust cytotoxicity and tumoricidal activity in PDAC, underscoring the translational potential for this novel class of antigens. Collectively, we have shown that cryptic epitopes, arising from aberrant translation in PDAC, represent a particularly promising class of antigens for next generation immune-based therapies. Citation Format: Zackery A Ely, Zachary J Kulstad, Gurcan Gunaydin, Sudarsana Addepalli, Eva K Verzani, Jennifer G Abelin, Marta Casarrubios, Karl R Clauser, Xilin Wang, Isabelle Lippincott, Cedric Louvet, Tom Schmitt, Miles Agus, Kevin S Kapner, Connor J Hennessey, James Cleary, Sine R Hadrup, Susan Klaeger, Jennifer Su, Alex M Jaeger, Brian M Wolpin, Srivatsan Raghavan, Eric Smith, Philip D Greenberg, Andrew J Aguirre, Steven A Carr, Tyler Jacks, William A Freed-Pastor. Discovery and therapeutic potential of novel cryptic peptides in pancreatic cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research; 2024 Sep 15-18; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl_2):Abstract nr PR-07.

Abstract We have been exploring in preclinical models and clinical trials methods to reproducibly provide therapeutic T-cell responses by transfer of genetically engineered T cells. Our largest clinical experience has been in treating human acute myelogenous leukemia (AML), in which we have utilized a high-affinity TCR specific for WT1, a protein associated with promoting leukemic transformation that is overexpressed in human leukemic stem cells, to genetically engineer CD8 T cells. We recently reported a study (Chapuis et al., Nat Med 2019) in which we treated leukemia patients at high risk of relapse (after hematopoietic cell transplant) that demonstrated all treated patients remain alive and relapse free at a median of 48 months, compared to a relapse rate of ~35% in the concurrent matched cohort (p<0.01). We have also been developing strategies to translate insights and technologies from this study to treatment of solid tumors. In a preclinical genetically engineered mouse model (KPC mice) of pancreatic cancer that faithfully replicates most aspects of human disease, we demonstrated (Stromnes et al., Cancer Cell 2015) that CD8 T cells engineered with a high-affinity TCR specific for mesothelin (Msln) can infiltrate pancreatic tumors, mediate antitumor activity, and provide therapeutic benefit. However, since the T cells are ultimately rendered dysfunctional in the tumor microenvironment (TME), prolonging survival has required repeated infusions of T cells to sustain antitumor activity. We have now isolated and validated a human high-affinity TCR specific for Msln for use in a planned clinical trial modeled after the approach successful in KPC mice. However, we would like to both enhance and sustain antitumor activity without requiring repeated infusions. In-depth analyses of the T cells, tumors, and the TME in treated KPC mice have illuminated strategies to potentially overcome the obstacles to tumor eradication, and we have been exploring molecular engineering approaches to achieve this. One approach has been to create synthetic immunomodulatory fusion proteins (IFPs) that have an ectodomain composed of the receptor for an inhibitory ligand encountered in the TME but, rather than the natural cytoplasmic tail that would deliver an inhibitory signal, the receptor has the tail of a costimulatory receptor and delivers an activation signal. Expression of such IFPs takes advantage of the inhibitory ligands commonly encountered in the TME by T cells by co-opting potential inhibitory signals and has resulted in enhanced T cell function, persistence/survival, and antitumor activity. Another major obstacle to sustained therapeutic activity appears to be the limited access in the TME to nutrients that effector T cells can utilize as an energy source. Analysis of the metabolites present in the TME and the transcriptional program in T cells has provided insights into genetic modifications that can be made to allow T cells to survive and function in the metabolically hostile TME. These and related studies will be discussed. Citation Format: Philip D. Greenberg, Kristin G. Anderson, Dan Egan, Sunil R. Hingorani, Luigi Nezi, Teresa Manzo, Shannon K. Oda, Kelly G. Paulson, Rachel Perret, Leah Schmidt, Tom M. Schmitt, Ingunn M. Stromnes, Aude G. Chapuis. Targeting pancreatic cancer with TCR-engineered T cells [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2019 Sept 6-9; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2019;79(24 Suppl):Abstract nr I11.

T cells specific for neoantigens encoded by mutated genes in cancers are increasingly recognized as mediators of tumor destruction after immune checkpoint inhibitor therapy or adoptive cell transfer. Unfortunately, most neoantigens result from random mutations and are patient specific, and some cancers contain few mutations to serve as potential antigens. We describe a patient with stage IV acral melanoma who achieved a complete response following adoptive transfer of tumor-infiltrating lymphocytes (TILs). Tumor exome sequencing surprisingly revealed fewer than 30 nonsynonymous somatic mutations, including oncogenic BRAFV600E. Analysis of the specificity of TILs identified rare CD4+ T cells specific for BRAFV600E and diverse CD8+ T cells reactive to nonmutated self-antigens. These specificities increased in blood after TIL transfer and persisted long-term, suggesting they contributed to the effective antitumor immune response. Gene transfer of the BRAFV600E-specific T cell receptor (TCR) conferred recognition of class II MHC-positive cells expressing the BRAF mutation. Therapy with TCR-engineered BRAFV600E-specific CD4+ T cells may have direct antitumor effects and augment CD8+ T cell responses to self- and/or mutated tumor antigens in patients with BRAF-mutated cancers.

Unlike cancer vaccines and immune modulators such as checkpoint inhibitors that seek to harness patient immune responses, adoptive therapy with genetically engineered T cells seeks to create responses that don’t exist in the patient’s immune system. Molecular technologies now make it feasible to not only create T cells with specificity for the tumor by introduction of a selected antigen-specific receptor, but also with qualities not naturally found, including improved function and resistance to immunosuppression. We have been exploring in preclinical models and clinical trials methods to reproducibly provide therapeutic T cell responses by transfer of genetically engineered T cells. For human acute myelogenous leukemia (AML), we have pursued targeting WT1, a gene overexpressed in human leukemic stem cells that is associated with promoting leukemic transformation. Preclinical studies performed in a mouse model demonstrated that CD8 T cells expressing a high affinity TCR specific for this oncogene can be safely administered, with no evidence of toxicity to the normal tissues known to express low but detectable levels of WT1. We have advanced this approach to a clinical trial in leukemia patients with poor prognostic factors that place them at high risk of relapse after hematopoietic cell transplant (HCT), using a high-affinity human TCR specific for WT1 to transduce CD8 cells and reproducibly create high-avidity T cells that recognize leukemic cells. Our clinical results demonstrate that such T cells can prevent leukemic relapse and sustain long-term remissions, and can mediate antileukemic activity in patients who have relapsed. This therapy is now being tested in AML patients who have minimal residual disease after induction therapy and are not candidates for HCT, as well as in solid tumors that similarly overexpress WT1. Unfortunately, there are substantive obstacles in targeting established tumors that can preclude even a T cell expressing a high-affinity TCR from being effective. These impediments include the development of T cell dysfunction, particularly within the microenvironment of solid tumors, and we are using genetically engineered mouse models to elucidate the cellular and molecular pathways that need to be modulated to achieve meaningful therapeutic benefit in a variety of hematologic and solid tumor settings, including pancreatic and ovarian cancer. Our preclinical therapy studies reveal promising antitumor activity, but demonstrate that repeated infusions of functional T cells are required to sustain a therapeutic response in the context of the immunosuppressive tumor microenvironment, and we are engineering T cells to overcome these inhibitory signals and enhance efficacy. In place of current strategies that disrupt inhibitory pathways by systemic administration of blocking mAbs, which globally disrupt immune regulation and thus can have significant toxicity to the host, we are creating synthetic immunomodulatory fusion proteins that take advantage of the expression of inhibitory ligands by tumors by still binding the inhibitory ligand but alternatively delivering a costimulatory rather than inhibitory signal. Additionally, as the antitumor activity of CD8 T cells is enhanced by a concurrent CD4 T cell response, we are engineering CD4 T cells as well as CD8 T cells to create an orchestrated antitumor response. The results suggest that cancer therapy with engineered T cells can provide effective antitumor responses and will likely find an increasing role in the treatment of human cancers. Citation Format: Philip D. Greenberg, Kristin G. Anderson, Dan Egan, Sunil R. Hingorani, Shannon K. Oda, Rachel Perret, Tom M. Schmitt, Ingunn M. Stromnes, Leah Schmidt, Aude G. Chapuis. Engineering T cells to eradicate tumors in the age of synthetic biology [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2017 Oct 1-4; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2018;6(9 Suppl):Abstract nr IA02.

Effective cellular therapy for human malignancies requires first identifying and validating an appropriate antigenic target, and then establishing in each patient a tumor-reactive T cell response of high avidity and high magnitude that is not only safe but can infiltrate and retain function in the tumor microenvironment. We have used molecular expression profiling to detect antigens selectively or markedly over-expressed by tumors, and then used these antigens as stimuli to generate T cells from normal repertoires. We have developed a high throughput technology to identify those T cells that express high affinity TCRs, and to then isolate from these T cells the TCR genes, place them either directly or after affinity enhancement into shuttle vectors, and use these reagents to create recipient T cells with high avidity for tumor targets that can be administered in vivo. We have utilized in silico, in vitro, and preclinical mouse models to assess the safety and potential efficacy of T cells expressing such TCRs. We are currently pursuing targeting of 3 antigens that are expressed in both murine and human tumors and are pro-oncogenic, contributing to the malignant phenotype. This includes ongoing clinical trials that will be discussed in acute myelogenous leukemia and in non-small cell lung cancer or mesothelioma targeting WT1 with T cells transduced to express a high affinity TCR specific for WT1, as well as trials being designed to target Mesothelin (MSLN) with T cells transduced to express a high affinity TCR specific for MSLN in pancreatic and ovarian cancer that are anticipated to begin within 8-12 months. Our clinical results in treatment of AML and preclinical results in mouse models of pancreatic and ovarian cancer appear very promising. However, the composite clinical data, as well as the results in the preclinical mouse models that drive our clinical trials, demonstrate substantial obstacles to sustaining T cell function in vivo after transfer, particularly in the context of solid tumors. Engineered T cells with specificity for tumor antigens appear to have the capacity to infiltrate and accumulate in solid tumors, and to initially mediate anti-tumor activity, but frequently become dysfunctional in the tumor microenvironment. Studies are being pursued to identify the critical obstacles to maintaining T cell function and achieving more reproducible tumor eradication, including modulating the tumor microenvironment and engineering T cells to express immunomodulatory fusion proteins (IFP) that recognize ligands for inhibitory signals but deliver an activation/costimulatory signal. Our data suggest that engineering T cells to acquire novel properties not naturally found in unmanipulated T cells has the potential to create effective therapies for human cancers. Citation Format: Philip D. Greenberg, Aude Chapuis, Dan Egan, Ingunn Stromnes, Sunil Hingorani, Shannon Oda, Rachel Perret, Kristin Anderson, Tom Schmitt. Building better T cells for targeting and eliminating tumors. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Targeting the Vulnerabilities of Cancer; May 16-19, 2016; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(1_Suppl):Abstract nr IA18.

Lentiviral vectors (LVs) pseudotyped with vesicular stomatitis virus envelope glycoprotein (VSV-G) have demonstrated great promise in gene therapy trials employing hematopoietic stem cell and T-cells. The VSV-G envelope confers broad tropism and stability to the vector but is toxic when constitutively expressed, which has impeded efforts to generate stable producer cell lines. We previously showed that cocal pseudotyped LVs offer an excellent alternative to VSV-G vectors because of their broad tropism and resistance to human serum inactivation. In this study, we demonstrate that cocal LVs transduce CD34(+) and CD4(+) T-cells more efficiently than VSV-G LVs and share the same receptor(s) for cell entry. 293T-cells stably expressing the cocal envelope produced significantly higher LV titers than VSV-G expressing cells. We developed cocal pseudotyped, third-generation, self-inactivating LV producer cell lines for a GFP reporter and for a WT1 tumor-specific T-cell receptor, which achieved concentrated titers above 10(8) IU/ml and were successfully adapted for growth in suspension, serum-free culture. The resulting LVs were at least as effective as standard LVs in transducing CD34(+) and CD4(+) T-cells. Our stable cocal LV producer cell lines should facilitate the production of large-scale, high titer clinical grade vectors.

Lentiviral vectors (LVs) are routinely used for stable gene transfer and have demonstrated great promise in hematopoietic stem cell gene therapy and also immunotherapy using genetically modified T cells. LVs are commonly pseudotyped with vesicular stomatitis virus envelope glycoprotein (VSV-G), which confers broad tropism to the vector and allows for vector concentration by centrifugation. However, the use of VSV-G has several limitations, such as susceptibility to inactivation by human serum complement making it unsuitable for in vivo delivery. Furthermore, VSV-G is toxic when constitutively expressed, which has impeded efforts to generate stable producer cell lines. In this study, we first validate the use of cocal vesiculovirus envelope to pseudotype LVs by demonstrating that cocal LVs transduce hematopoietic stem cells and CD4+ T cells more efficiently than VSV-G LVs. We also provide evidence that cocal and VSV-G envelopes use the same receptor for cell entry. We then describe the development of two high-titer, cocal-pseudotyped, LV producer cell lines for a GFP reporter and for a WT1 tumor-specific T cell receptor (TCR). The different 3rd generation lentiviral helper genes were sequentially introduced in HEK293T cells by co-transfection with plasmids encoding antibiotic resistance genes followed by selection to allow for stable protein expression. Cells expressing the cocal envelope produced over 10-times more infectious LV particles as compared to VSV-G expressing cells. High-titer cocal producer cells were isolated by screening for best single clones, which were capable of generating concentrated titers above 108 infectious units per mL. We found that these producer cells were stable after serial passages for over 3 months, with no drop in titer detected over time. The resulting GFP and WT1-TCR vectors performed at least as well as identical vectors made with our standard transient transfection protocol for the transduction of CD34+ and CD4+ T cells, respectively. Cocal LV producer cells were also adapted for growth in suspension, serum-free culture, which will facilitate efforts for the scaling up of vector production. In summary, we have successfully developed two independent LV producer cells lines with clinically usable titers. The broad applicability of our cocal packaging cell line offers a promising tool toward the generation of large-scale, clinical grade LV.

Abstract We have entered a new and exciting era in cancer therapy, in which immunotherapeutic strategies are achieving unprecedented successes and are increasingly becoming incorporated into standard of care regimens. Checkpoint blockade is dependent on inducing and/or reactivating or sustaining responses to the tumor by T cells already in the patient. Similarly, vaccines attempt to generate and/or expand responses of T cells naturally present in the normal repertoire. However, these strategies require that functional tumor-reactive T cells exist in the patient's repertoire and that the method pursued can harness those T cells to create a potent response that will function in the tumor microenvironment, which limits the settings in which these approaches will prove effective. Adoptive T cell therapy, in which patient T cells can be expanded to large numbers ex vivo before infusion, provides a means to bypass or overcome these obstacles, particularly with the advent of genetic engineering that now makes it possible to create T cells not only with specificity for the tumor but also with qualities not naturally found, including improved function and resistance to immunosuppression. We have been exploring in preclinical models and clinical trials methods to reproducibly provide therapeutic T cell responses by transfer of genetically engineered T cells. The first issue is to identify tumor antigens that can be safely, effectively, and reproducibly targeted. We have used analyses of differential gene expression to search for antigenic targets that are either uniquely expressed in a tumor or are differentially expressed at high levels in the tumor with much lower and limited expression in normal tissues, and that preferentially are associated with the malignant phenotype to reduce the risk of antigen loss by the tumor. In our search for targets in acute myelogenous leukemia (AML), we found that WT1, a gene known to be associated with promoting leukemic transformation, is expressed in comparative abundance in human leukemic stem cells. The next step is to generate T cells specific for the target antigen that can recognize and eliminate malignant cells expressing the antigen. Extensive screening of normal human repertoires revealed a high affinity TCR specific for WT1 that can recognize leukemic cells, and that could be inserted into CD8 T cells to reproducibly produce high avidity T cells for use in therapy. Preclinical studies performed in a mouse model demonstrated that CD8 T cells specific for this oncogene expressing a high affinity TCR can be safely administered, with no evidence of toxicity to the normal tissues known to express low but detectable levels of WT1. We have advanced this approach targeting WT1 to an initial clinical trial in leukemia patients with poor prognostic factors that make them at high risk of relapse after hematopoietic cell transplant (HCT). The Vα and Vβ genes of the human WT-1 specific TCR were codon optimized to enhance expression, modified by a point mutation in each chain to create an interchain disulfide bond that minimizes the potential problem of mispairing of the introduced TCR chains with the endogenous TCR chains, and inserted these TCR genes into a lentiviral vector. Preliminary results of this trial, which has provided evidence that such T cells can prevent leukemic relapse and sustain long-term remissions, will be discussed. This therapy is now being advanced for use in AML patients who are not HCT candidates. We have also now initiated additional trials with this TCR for treatment of patients with non-small cell lung cancer (NSCLC) or mesothelioma, as WT1 is commonly overexpressed in NSCLC as well as many other malignancies. For many candidate target antigens that are also normal self-antigens, isolation of high affinity TCRs may not be readily achieved from normal repertoires. However, it is now feasible to engineer TCRs that have higher affinities than normally exist for their antigen target. We have developed strategies to enhance the affinity of isolated TCRs with retention of specificity, including saturation mutagenesis of CDR3 regions and an in vitro thymic selection system that allows for capture of a more diverse set of high affinity specific TCRs during TCR gene rearrangement. These approaches induce modifications to the TCR region that predominantly makes contacts with the peptide epitope rather than MHC, which is necessary to minimize the risk of off-target toxicity from promiscuous peptide/MHC recognition. However, it remains essential that such modified TCRs do not induce unanticipated tissue damage, and we are using bioinformatics, functional screening, and modeling in the mouse to uncover any potential for off-target toxicity. Unfortunately, providing a high avidity T cell response does not necessarily result in tumor eradication, as there are other substantive obstacles that can preclude even a T cell expressing a high affinity TCR from being effective. These impediments include the development of T cell dysfunction, particularly within the microenvironment of solid tumors, and we are using genetically engineered mouse models to elucidate the cellular and molecular pathways that need to be modulated to achieve meaningful therapeutic benefit in a variety of solid tumor settings, including pancreatic and ovarian cancer. Our preclinical therapy studies, particularly in a pancreatic ductal adenocarcinoma (PDA) model, already appear very promising, as we have demonstrated that T cells expressing a high affinity TCR targeting a tumor antigen expressed by PDA cells can infiltrate the tumor, mediate tumor lysis, modify the tumor stroma, and provide therapeutic benefit. We have now identified high affinity human TCRs specific for this tumor antigen, and plan to use the insights derived from these studies to initiate within the next year clinical trials in human pancreatic and ovarian cancers. The genetically-engineered mouse models of spontaneously developing tumors we are using, which recapitulate many aspects of the analogous human cancer, are also making it possible to assess strategies to improve the efficacy of T cell therapy. These models have helped elucidate the importance of not only cell extrinsic mechanisms of regulation and dysfunction that render T cells unresponsive, particularly via inhibitory cells commonly present in the tumor microenvironment that interfere with an effector response, such as the accumulation of regulatory CD4 T cells (Treg), myeloid derived suppressor cells (MDSC), and tumor-associated macrophages (TAM), but also the cell intrinsic mechanisms that derive in large part from persistent stimulation by the tumor antigen and ultimately can render T cells progressively dysfunctional, leading to epigenetic modifications that eventuallly result in non-responsive cells that cannot be readily rescued. These cumulative mechanisms highlight the difficulties eliciting and/or sustaining responses to tumor antigens. Strategies to disrupt inhibitory pathways by systemic administration of mAbs or cytokines are currently being pursued clinically, but such reagents globally disrupt inhibitory pathways and thus can have significant toxicity to the host. Therefore, we are evaluating strategies to sustain function and anti-tumor activity by genetically modifying T cells to enhance function and to be resistant to obstacles that prevent tumor eradication. As different tumor types exhibit unique characteristics and are capable of engaging distinct inhibitory pathways, improved understanding of the immunobiology of the tumor type to be treated will likely prove essential for designing effective therapies. However, the relatively straightforward means to use synthetic biology to genetically engineer T cells to acquire novel capacities to overcome inhibitory signals and function in the tumor microenvironment suggests that cancer therapy with engineered T cells will likely find an increasing role in the treatment of human cancers. Citation Format: Philip D. Greenberg, Kristin G. Anderson, Dan Egan, Sunil R. Hingorani, Shannon K. Oda, Rachel Perret, Andrea Schietinger, Tom M. Schmitt, Ingunn M. Stromnes, Alec Wilkens, Aude G. Chapuis. Engineering T cell responses to tumors: Taking the immune system where no responses have gone before [abstract]. In: Proceedings of the Second CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; 2016 Sept 25-28; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2016;4(11 Suppl):Abstract nr IA01.

Effective cellular therapy for human malignancies requires first identifying and validating an appropriate antigenic target, and then establishing in each patient a tumor-reactive T cell response of high avidity and high magnitude that is safe and can infiltrate and retain function in the tumor microenvironment. We have been exploring in preclinical models and clinical trials methods to reproducibly provide such responses by transfer of genetically engineered T cells that acquire target specificity by virtue of an introduced high affinity TCR. To identify candidate antigens in leukema, we examined purified human leukemic stem cells for over-expression of genes based on comparisons to purified human hematopoietic stem cells as well as normal somatic tissues. Our analysis revealed that WT1, a gene known to be associated with promoting leukemic transformation, is expressed in comparative abundance in human leukemic stem cells. Preclinical studies were then performed in a mouse model, and revealed that CD8 T cells specific for this oncogene with even higher avidity than can be detected in normal repertoires could be safely administered, with no evidence of toxicity to the normal tissues known to express low but detectable levels of WT1. For our initial clinical trial, poor prognosis leukemia patients who relapsed after hematopoietic cell transplant (HCT) were treated with transfer of WT1-specific CD8 T cells clones isolated and expanded in vitro from the HCT donor. This study demonstrated that such T cells were safe, mediated in vivo anti-leukemic activity, and were associated with maintenance of long-term remissions in some patients, but generating sufficient numbers of WT1-specific CD8 T cells with high avidity for the target in each patient represented a substantive problem. Therefore, to create a more predictably effective standardized reagent for treatment of patients with a tumor that expresses the target antigen and shares the associated MHC restricting allele, we pursued methods to genetically engineer patient T cells to acquire high avidity for the tumor target. This requires identifying a high affinity TCR and producing a vector that can achieve high-level expression of the genes encoding the Vα and Vβ genes of a TCR demonstrated to have high affinity for the target epitope. Therefore, we screened a large number of normal repertoires for the presence of high avidity WT1-specific CD8 T cells, and selected the T cell clone expressing the highest affinity TCR. We then incorporated changes in the TCR genes such as codon optimization to enhance expression, and introduced a point mutation in each chain to create a disulfide bond that minimizes the potential problem of mispairing of the introduced TCR chains with the endogenous TCR chains. We have now have now initiated a trial in which this high affinity, WT1-specific, HLA-A2-restricted TCR is being introduced into patient CD8 T cells with a lentiviral vector and the transduced cells are being infused into the patient. The early results from this trial appear promising in terms of both evidence of antileukemic activity and the capacity for the transferred cells to persist in patients, and we plan to begin very shortly another trial in patients with non-small cell lung cancer (NSCLC) utilizing this same TCR, as WT1 is also commonly overexpressed in NSCLC as well as many other malignancies. For many candidate target antigens that are also normal self-antigens, isolating high affinity TCRs may not be readily achieved from normal repertoires. Therefore, we have developed strategies to enhance the affinity of isolated TCRs with retention of specificity, including saturation mutagenesis of CDR3 regions and an in vitro thymic selection system that allows for capture of a more diverse set of high affinity specific TCR genes during TCR gene rearrangement. These approaches induce modifications to the TCR region that predominantly makes contacts with the peptide epitope rather than MHC, which is necessary to minimize the risk of off-target toxicity from promiscuous peptide/MHC recognition. However, it remains essential that such modified TCRs do not induce unanticipated tissue damage, and we are using bioinformatics as well as modeling in the mouse to uncover any potential for off-target toxicity. Unfortunately, providing a high avidity T cell response does not necessarily result in tumor eradication, as there are other substantive obstacles that can preclude even a T cell expressing a high affinity TCR from being effective. These impediments include the development of T cell dysfunction, particularly within the microenvironment of solid tumors, and we are using genetically engineered mouse models to elucidate the cellular and molecular pathways that need to be modulated to achieve meaningful therapeutic benefit in a variety of solid tumor settings, including pancreatic and ovarian cancer. Our preclinical therapy studies, particularly in a pancreatic ductal adenocarcinoma (PDA) model, already appear very promising, as we have demonstrated that T cells expressing a high affinity TCR targeting a tumor antigen expressed by PDA cells can infiltrate the tumor, mediate tumor lysis, modify the tumor stroma, and provide therapeutic benefit. We have already identified high affinity human TCRs specific for this tumor antigen, and plan to use the insights derived from these studies to initiate within the next 1-2 years clinical trials in human pancreatic and ovarian cancers. The genetically-engineered mouse models of spontaneously developing tumors we are using, which recapitulate many aspects of the analogous human cancer, are also making it possible to assess strategies to improve the efficacy of T cell therapy. These models have helped elucidate the importance of not only cell extrinsic mechanisms of regulation and dysfunction that render T cells unresponsive, particularly via inhibitory cells commonly present in the tumor microenvironment that interfere with an effector response such as the accumulation of regulatory CD4 T cells (Treg), myeloid derived suppressor cells (MDSC), and tumor-associated macrophages (TAM), but also the cell intrinsic mechanisms that derive in large part from persistent stimulation by the tumor antigen and ultimately can render T cells progressively dysfunctional, leading to epigenetic modifications that eventually result in non-responsive cells that cannot be readily rescued. These cumulative mechanisms highlight the difficulties eliciting and/or sustaining responses to tumor antigens. Strategies to disrupt inhibitory pathways by systemic administration of mAbs or cytokines are currently being pursued clinically, but such reagents globally disrupt inhibitory pathways which can have significant toxicity to the host. Therefore, we are evaluating strategies to sustain function and anti-tumor activity by genetically modifying T cells to enhance function and to be resistant to obstacles that prevent tumor eradication. As different tumor types exhibit unique characteristics and are capable of engaging distinct inhibitory pathways, improved understanding of the immunobiology of the tumor type to be treated will likely prove essential for designing effective therapies. However, the relatively straightforward means to use synthetic biology to genetically engineer T cells to acquire novel capacities to overcome inhibitory signals and function in the tumor microenvironment suggests that cancer therapy with engineered T cells will likely find an increasing role in the treatment of human cancers. Citation Format: Philip D. Greenberg, Tom M. Schmitt, Andrea Schietinger, Ingunn M. Stromnes, Sunil R. Hingorani, Shannon K. Oda, Rachel Perret, Kristin G. Anderson, Merav Bar, Aude G. Chapuis. Employing TCRs in engineered T cells to develop therapeutic reagents for effectively targeting malignancies. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr SY31-03. doi:10.1158/1538-7445.AM2015-SY31-03

Effective cellular therapy for human malignancies requires first identifying and validating an appropriate antigenic target, and then establishing in each patient a tumor-reactive T cell response of high avidity and high magnitude that is safe and can infiltrate and retain function in the tumor microenvironment. We have been exploring in preclinical models and clinical trials methods to reproducibly provide such responses by transfer of genetically engineered T cells that acquire target specificity by virtue of an introduced high affinity TCR. To identify candidate antigens in leukema, we examined purified human leukemic stem cells for over-expression of genes based on comparisons to purified human hematopoietic stem cells as well as normal somatic tissues. Our analysis revealed that WT1, a gene known to be associated with promoting leukemic transformation, is expressed in comparative abundance in human leukemic stem cells. Preclinical studies were then performed in a mouse model, and revealed that CD8 T cells specific for this oncogene with even higher avidity than can be detected in normal repertoires could be safely administered, with no evidence of toxicity to the normal tissues known to express low but detectable levels of WT1. For our initial clinical trial, poor prognosis leukemia patients who relapsed after hematopoietic cell transplant (HCT) were treated with transfer of WT1-specific CD8 T cells clones isolated and expanded in vitro from the HCT donor. This study demonstrated that such T cells were safe, mediated in vivo anti-leukemic activity, and were associated with maintenance of long-term remissions in some patients, but generating sufficient numbers of WT1-specific CD8 T cells with high avidity for the target in each patient represented a substantive problem. Therefore, to create a more predictably effective standardized reagent for treatment of patients with a tumor that expresses the target antigen and shares the associated MHC restricting allele, we pursued methods to genetically engineer patient T cells to acquire high avidity for the tumor target. This requires identifying a high affinity TCR and producing a vector that can achieve high-level expression of the genes encoding the Vα and Vβ genes of a TCR demonstrated to have high affinity for the target epitope. Therefore, we screened a large number of normal repertoires for the presence of high avidity WT1-specific CD8 T cells, and selected the T cell clone expressing the highest affinity TCR. We then incorporated changes in the TCR genes such as codon optimization to enhance expression, and introduced a point mutation in each chain to create a disulfide bond that minimizes the potential problem of mispairing of the introduced TCR chains with the endogenous TCR chains. We have now have now initiated a trial in which this high affinity, WT1-specific, HLA-A2-restricted TCR is being introduced into patient CD8 T cells with a lentiviral vector and the transduced cells are being infused into the patient. The early results from this trial appear promising in terms of both evidence of antileukemic activity and the capacity for the transferred cells to persist in patients, and we plan to begin very shortly another trial in patients with non-small cell lung cancer (NSCLC) utilizing this same TCR, as WT1 is also commonly overexpressed in NSCLC as well as many other malignancies. For many candidate target antigens that are also normal self-antigens, isolating high affinity TCRs may not be readily achieved from normal repertoires. Therefore, we have developed strategies to enhance the affinity of isolated TCRs with retention of specificity, including saturation mutagenesis of CDR3 regions and an in vitro thymic selection system that allows for capture of a more diverse set of high affinity specific TCR genes during TCR gene rearrangement. These approaches induce modifications to the TCR region that predominantly makes contacts with the peptide epitope rather than MHC, which is necessary to minimize the risk of off-target toxicity from promiscuous peptide/MHC recognition. However, it remains essential that such modified TCRs do not induce unanticipated tissue damage, and we are using bioinformatics as well as modeling in the mouse to uncover any potential for off-target toxicity. Unfortunately, providing a high avidity T cell response does not necessarily result in tumor eradication, as there are other substantive obstacles that can preclude even a T cell expressing a high affinity TCR from being effective. These impediments include the development of T cell dysfunction, particularly within the microenvironment of solid tumors, and we are using genetically engineered mouse models to elucidate the cellular and molecular pathways that need to be modulated to achieve meaningful therapeutic benefit in a variety of solid tumor settings, including pancreatic and ovarian cancer. Our preclinical therapy studies, particularly in a pancreatic ductal adenocarcinoma (PDA) model, already appear very promising, as we have demonstrated that T cells expressing a high affinity TCR targeting a tumor antigen expressed by PDA cells can infiltrate the tumor, mediate tumor lysis, modify the tumor stroma, and provide therapeutic benefit. We have already identified high affinity human TCRs specific for this tumor antigen, and plan to use the insights derived from these studies to initiate within the next 1-2 years clinical trials in human pancreatic and ovarian cancers. The genetically-engineered mouse models of spontaneously developing tumors we are using, which recapitulate many aspects of the analogous human cancer, are also making it possible to assess strategies to improve the efficacy of T cell therapy. These models have helped elucidate the importance of not only cell extrinsic mechanisms of regulation and dysfunction that render T cells unresponsive, particularly via inhibitory cells commonly present in the tumor microenvironment that interfere with an effector response such as the accumulation of regulatory CD4 T cells (Treg), myeloid derived suppressor cells (MDSC), and tumor-associated macrophages (TAM), but also the cell intrinsic mechanisms that derive in large part from persistent stimulation by the tumor antigen and ultimately can render T cells progressively dysfunctional, leading to epigenetic modifications that eventually result in non-responsive cells that cannot be readily rescued. These cumulative mechanisms highlight the difficulties eliciting and/or sustaining responses to tumor antigens. Strategies to disrupt inhibitory pathways by systemic administration of mAbs or cytokines are currently being pursued clinically, but such reagents globally disrupt inhibitory pathways which can have significant toxicity to the host. Therefore, we are evaluating strategies to sustain function and anti-tumor activity by genetically modifying T cells to enhance function and to be resistant to obstacles that prevent tumor eradication. As different tumor types exhibit unique characteristics and are capable of engaging distinct inhibitory pathways, improved understanding of the immunobiology of the tumor type to be treated will likely prove essential for designing effective therapies. However, the relatively straightforward means to use synthetic biology to genetically engineer T cells to acquire novel capacities to overcome inhibitory signals and function in the tumor microenvironment suggests that cancer therapy with engineered T cells will likely find an increasing role in the treatment of human cancers. Citation Format: Philip D. Greenberg, Tom M. Schmitt, Andrea Schietinger, Ingunn M. Stromnes, Sunil R. Hingorani, Shannon K. Oda, Rachel Perret, Kristin G. Anderson, Merav Bar, Aude G. Chapuis. Employing TCRs in engineered T cells to develop therapeutic reagents for effectively targeting malignancies. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr SY31-03. doi:10.1158/1538-7445.AM2015-SY31-03

Abstract Modulating T cell immunity to treat human malignancies is showing increasing promise, but substantive obstacles remain. Reproducibly effective therapy will require many factors be in place, including an appropriate antigenic target, a high avidity and high magnitude T cell response, and the T cells having the ability to infiltrate and retain function in the tumor microenvironment. We have been systematically investigating strategies to address these issues. To identify targetable leukemia antigens, we compared gene expression profiles in purified leukemic stem cells with profiles in normal hematopoietic stem cells and other somatic cells, and identified two promising targets that are associated with the leukemic phenotype, WT1 and Cyclin A1, and to which CD8 T cells that lyse leukemic cells can be generated. Generating sufficient numbers of specific T cells with high avidity for the target in each patient is a substantive problem. We will discuss a just completed trial targeting WT1 in leukemia patients, which highlights the potential benefit of providing potent T cell responses to this pro-oncogenic protein. For this trial, we generated panels of WT1-specific CD8 T cells clones for each patient and then selected and expanded for adoptive therapy the highest avidity clone isolated. However, this approach is limited by the responses elicited for each patient, who often have compromised repertoires, and could be overcome by identifying the T cell receptor (TCR) gene from a defined high avidity leukemia-reactive T cell clone that can be introduced into large numbers of patient T cells to create a standardized reagent for treatment. However, ultimately the avidity of transduced T cells used for therapy is limited by the affinity of the introduced TCR, and high affinity TCRs for tumor antigens that are also normal self-antigens may not be readily identified in normal repertoires. Our lab has developed in collaboration with David Kranz' lab methods to mutate/alter the CDR3 regions of the isolated antigen-specific TCR chains prior to introduction into recipient T cells to improve the affinity for the target antigen, as well as methods to interrogate the full repertoire of β-chain rearrangements capable of pairing with a defined TCR α-chain before negative selection. Strategies to evaluate the in vivo activity and safety of such TCRs in relevant mouse models will be described. Unfortunately, providing a high avidity T cell response does not necessarily result in tumor eradication. Major obstacles include the development of tolerance/anergy and/or exhaustion/dysfunction in tumor-reactive T cells, particularly within the tumor microenvironment. We have explored these issues in T cell therapy models in mice with leukemia or that “spontaneously” develop solid tumors as a consequence of regulated tissue-specific expression of an oncogene. These studies highlight the difficulties sustaining responses to tumor antigens that are self-proteins and the inhibitory pathways that are commonly operative within the tumor microenvironment, and have provided insights into how to potentially sustain activity by selecting or genetically modifying T cells to be resistant to obstacles that prevent tumor eradication. However, different tumor types can engage distinct pathways and have unique characteristics, and thus understanding the immunobiology of the tumor to be treated will likely be essential for designing effective therapies. Citation Format: Philip D. Greenberg, Sebastian Ochsenreither, Tom Schmitt, David Aggen, David Kranz, Matthias Wolfl, Jurgen Kuball, Ravi Majeti, Irv Weissman, Ingunn Stromnes, Andrea Schietinger, Gunnar Ragnarsson, Cassian Yee, Merav Bar, Aude Chapuis. T cells vs. tumor cells: Arming/deploying T cells for a successful battle. [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology: Multidisciplinary Science Driving Basic and Clinical Advances; Dec 2-5, 2012; Miami, FL. Philadelphia (PA): AACR; Cancer Res 2013;73(1 Suppl):Abstract nr IA1.