A personalised approach for anti-GITR-based immunotherapy in pre-clinical models of pancreatic ductal adenocarcinoma

Abstract

Abstract Background Patient outcomes with pancreatic ductal adenocarcinoma (PDA) remain dismal, despite the introduction of combination chemotherapies. Response to therapy is hampered by tumour stroma, which is highly heterogeneous but poorly characterised. The clinical success of immunotherapy in certain cancer types justifies investigation in PDA using a personalised approach of selecting pre-clinical models. Methods Patient PDA samples (n = 70; International Cancer Genome Consortium) were subjected to hierarchical clustering using a gene set representing CD8 and regulatory T cells to identify sub-groups. Syngeneic mouse models recapitulating these sub-groups (cross-species analysis) were developed using cell lines from genetically engineered mouse (GEM) models. Models bearing subcutaneous tumours were treated with anti-GITR agonist (DTA-1; 10mg/kg) or isotype (IgG2b) and tissues were collected for 12-color fluorescence activated cell sorting (FACS), NanoString and immunohistochemistry (IHC). Results Using clustering analysis, we identified four distinct patient PDA immune sub-groups with differential expression of T cell genes. These sub-groups showed differential expression of glucocorticoid-induced TNF receptor (GITR) and its associated genes. Hence, we hypothesised that these groups may respond differently to anti-GITR therapy in vivo. As expected, the three syngeneic models showed differential response to anti-GITR treatment. Model with increased GITR expression (>2-fold; NanoString and immunoblot) showed a significant (p = 0.02) survival benefit and a modest decrease in tumour burden upon anti-GITR treatment (n = 10) compared to the isotype control (n = 7) and the other two models. We observed anti-GITR mediated increase in GITR+ CD8+ cells and no change in GITR+ FOXP3+ cells (FACS, NanoString and IHC) along with increased gene expression of Pdcd1, Lag3, Spn and Itk. Conclusion Our models recapitulate the immunological gene expression profiles observed in patient PDA samples making them a highly useful resource to study personalized immunotherapies. This integrated study highlights the importance of patient selection to maximize therapy benefit and spare immune related adverse events. Legal entity responsible for the study The authors. Funding Bristol-Myers Squibb. Disclosure D. Cunningham: Research grant / Funding (self): Amgen; Research grant / Funding (self): Sanofi; Research grant / Funding (self): Merrimack; Research grant / Funding (self): AstraZeneca; Research grant / Funding (self): Celgene; Research grant / Funding (self): MedImmune; Research grant / Funding (self): Bayer; Research grant / Funding (self): 4SC; Research grant / Funding (self): Clovis; Research grant / Funding (self): Eli Lilly; Research grant / Funding (self): Janssen; Research grant / Funding (self): Merck. N. Starling: Research grant / Funding (self): AstraZeneca; Research grant / Funding (self): Bristol-Myers Squibb; Research grant / Funding (self): Pfizer; Travel / Accommodation / Expenses: AstraZeneca; Travel / Accommodation / Expenses: Bristol-Myers Squibb; Travel / Accommodation / Expenses: Eli Lilly; Travel / Accommodation / Expenses: Merck; Travel / Accommodation / Expenses: Roche; Honoraria (self): Eli Lilly; Honoraria (self): Merck Serono; Honoraria (self): MSD Oncology; Honoraria (self): Pierre Fabre; Advisory / Consultancy: Pfizer; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Servier. A. Sadanandam: Research grant / Funding (self): Bristol-Myers Squibb; Research grant / Funding (self): Merck KGaa; Research grant / Funding (self): Pierre Fabre. All other authors have declared no conflicts of interest.