Adoptive transfer of engineered T cells prolongs survival in preclinical ovarian cancer

From the Greenberg lab, Clinical Research Division, and the Vaccine and Infectious Disease Division 

Patients with high-grade serous ovarian cancer (HGSOC) are prone to relapse and low survival rates following current best treatment strategies. Engineered chimeric antigen receptor (CAR) T cells have been successful in treating blood cancers, but their use in solid tumors such as HGSOC requires further research. Mesothelin (MSLN), a glycoprotein that is overexpressed in tumor cells, is of interest as a target for engineered adoptive T cell therapy but its efficacy has not been validated. Additionally, the transplantable ID8 murine ovarian cancer cell line, which is transduced to overexpress vascular endothelial growth factor (ID8VEGF), is a promising model for studying HGSOC immunotherapy strategies, but it is unknown how closely ID8VEGF  mimics the immunosuppressive environment of human HGSOC. To address these immunotherapy questions in the context of HGSOC, Kristin Anderson, a scientist from the Greenberg lab (Clinical Research Division), along with colleagues from the Gottardo lab (Vaccine and Infectious Disease Division) investigated the use of this model in HGSOC research as well as the efficacy of CD8+ anti-MSLN T cells in both mouse and human tumors. They recently published their work in Cancer Immunology Research.

To determine whether human T cells transduced with MSLN-specific TCRs could recognize and kill tumor cells, the authors stimulated their previously developed MSLN-targeting T cells with cognate peptide and found that these cells could produce inflammatory cytokines and bind MSLN:MHC complexes, as well as kill MSLN-expressing HGSOC cells in culture. To further validate this system, the authors confirmed that MSLN is overexpressed in mouse ovarian tumors, just as it is in human HGSOC. The authors then evaluated if engineered TCR-T mouse cells could respond to TCR signals. Their previously engineered MSLN-specific murine CD8+ T cells could bind MSLN tetramers and produce inflammatory cytokines, demonstrating their ability to recognize and respond to their cognate antigen through TCR signaling. However, to confirm that engineered TCR-T cells could recognize endogenously expressed MSLN, MSLN-specific mouse T cells, along with control T cells specific for an irrelevant, non-MSLN antigen, were co-cultured with ID8VEGF cells. MSLN T cells, but not control T cells, proliferated in culture and lysed cells in an effector:target dose-dependent manner, while expressing low levels of inhibitory or exhaustion markers, which can negatively influence anti-tumor capacity of effector T cells. Together, these experiments demonstrated that MSLN is a viable immunotherapy target and that T cells designed to recognize MSLN were successful in finding and killing cancer cells.

The authors next sought to validate the murine ID8VEGF model as an appropriate tool to study human HGSOC by performing deep transcriptional profiling on mouse and human ovarian tumors. Predictably, species-specific gene expression differences were found between the two, but many KEGG pathways, as well as the majority of antigen-presentation and immunosuppressive pathways were similarly expressed between the two tumors. While the general shared transcriptional signature between mouse and human tumors is important, tumor cells often express specific immunosuppressive proteins to dampen host anti-tumor immune responses within the tumor microenvironment (TME), perpetuating their own growth. To more narrowly compare this aspect of the TME between species, the authors performed immunohistochemistry to visualize the presence of immunosuppressive cell subsets such as regulatory T cells, as well as inhibitory ligands such as PD-L1. They found that both models consist of similar TME makeups, further supporting ID8VEGF as a valid model to study immunotherapy strategies for human HGSOC. 

Treating mice with T cells engineered with anti-MSLN TCRs and subsequent Msln vaccine boostings prolongs prolongs survival in mice with ovarian tumors. Figure provided by Kristin Anderson.
Treating mice with T cells engineered with anti-MSLN TCRs and subsequent Msln vaccine boostings prolongs prolongs survival in mice with ovarian tumors. Figure provided by Kristin Anderson.

Moving this research in vivo, the authors next investigated the functional capacity of anti-tumor T cells to prolong survival in mice with ovarian cancer. They transferred T cells specific for MSLN, or T cells specific for an irrelevant antigen, into mice with established ID8VEGF ovarian tumors. They found that although MSLN-specific T cells, but not control cells, were enriched in the tumor, their presence did not increase survival. However, after testing several adoptive transfer scenarios, the authors found that when the mice were given repeated doses of MSLN-T cells and a T cell-stimulating “vaccination” with MSLN antigen, their survival was increased over untreated or mice or those treated with the control T cells. They used immunohistochemistry images of tumor sections to confirm that the mice that received anti-MSLN cells had increased levels of tumor cell death, suggesting that their survival advantage was due to anti-tumor activity, where engineered T cells were better able to target and kill tumor cells.

This work made important findings in HGSOC research as well as more generally in the field of cancer immunotherapy, beginning with demonstrating the similarities between human HGSOC and the murine ID8VEGF transplantable tumor model and validating its use in immunotherapy target research. This research also identifies important differences between the two models, which may explain discrepancies observed between pre-clinical mouse research and human clinical trials. Importantly, this research also “demonstrates that engineered adoptive T cell therapy can kill ovarian cancer and prolong survival using a model of advanced stage disease,” Anderson said. “Specifically, we have shown that mesothelin is a valid target antigen in high grade serous ovarian cancer. This work both sets the stage for initiating immunotherapy with TCR-engineered T cells in human ovarian cancer and has established a preclinical model useful for examining iterative strategies that address obstacles to efficacy that will likely be encountered in clinical trials.” Moving forward, Anderson said that their research will focus on both the ID8VEGF model itself as well as its implementation in designing HGSOC-specific T cells. “Our work has confirmed several major obstacles to adoptive T cell immunotherapy in ovarian cancer, and validated that the majority of these obstacles are present in the mouse model of disease as well,” Anderson explained. “Our future work will investigate strategies for overcoming these obstacles, including combination therapies and additional T cell engineering approaches.”

This work was supported by the Chromosome Metabolism and Cancer Training Grant Program, the Ovarian Cancer Research Alliance, the Solid Tumor Translational Research Award, the National Institutes of Health National Cancer Institute, the Leukemia & Lymphoma Society, Juno Therapeutics, and the Parker Institute for Cancer Immunotherapy. 

UW/Fred Hutch Cancer Consortium members Charles Drescher, Raphael Gottardo, and Phil Greenberg contributed to this work. 

Anderson KG, Voillet V, Bates BM, Chiu EY, Burnett MG, Garcia NM, Oda SK, Morse CB, Stromnes IM, Drescher CW, Gottardo R, Greenberg PD. 2019. Engineered adoptive T-cell therapy prolongs survival in a preclinical model of advanced-stage ovarian cancer. Cancer Immunology Research. 2019 Sep;7(9):1412-1425. doi: 10.1158/2326-6066.CIR-19-0258. Epub 2019 Jul 23.