Multimodal Mapping of the Tumor and Peripheral Blood Immune Landscape in Human Pancreatic Cancer

Pancreatic ductal adenocarcinoma (PDA) is characterized by an immune-suppressive tumor microenvironment that renders it largely refractory to immunotherapy. We implemented a multimodal analysis approach to elucidate the immune landscape in PDA. Using a combination of CyTOF, single-cell RNA sequencing, and multiplex immunohistochemistry on patient tumors, matched blood, and non-malignant samples, we uncovered a complex network of immune-suppressive cellular interactions. These experiments revealed heterogeneous expression of immune checkpoint receptors in individual patient's T cells and increased markers of CD8+ T cell dysfunction in advanced disease stage. Tumor-infiltrating CD8+ T cells had an increased proportion of cells expressing an exhausted expression profile that included upregulation of the immune checkpoint TIGIT, a finding that we validated at the protein level. Our findings point to a profound alteration of the immune landscape of tumors, and to patient-specific immune changes that should be taken into account as combination immunotherapy becomes available for pancreatic cancer.

Abstract 3442: Multimodal mapping of the immune landscape in human pancreatic cancer

Background: Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of cancer related death in the US. Unfortunately, recent clinical trials using immunotherapy targeting the highly immunosuppressive tumor microenvironment have showed disappointing results, as there is no method to predict which patients will respond to therapy. More recently, the development of single cell technology has allowed for in-depth profiling at the cellular level using small amounts of tissue, raising the potential to develop precision medicine tools at time of endoscopic fine needle biopsy.

Results: We performed single-cell RNA sequencing (scRNAseq) on endoscopic fine needle biopsies from 10 PDAC tumors at time of diagnostic endoscopic biopsy or 6 surgically resected tissues. We also sequenced 3 adjacent or normal pancreas tissues. We captured 8,521 cells from 3 surgical normal adjacent samples and 46,244 cells from PDAC tumors. Mapping of putative interactions between ligands and receptors demonstrated upregulation of key signaling pathways, including Hedgehog, NOTCH, and chemokine signaling within myeloid, epithelial, T, and NK cells. Differential expression analysis in cytotoxic CD8 T cells of PDA patients revealed increased expression of genes involved in T cell activation (GZMB, GZMA), exhaustion (GZMK, EOMES) as well as immune checkpoint pathway upregulation when compared to cytotoxic T cells in adjacent normal pancreatic tissue. Among the most significantly increased genes in CD8 T cells of PDAC tumor was the immune checkpoint TIGIT. Upon further analysis of the CD8 T cells, we found TIGIT was almost exclusively expressed in exhausted CD8 T cells, while other checkpoints such as PD-1 and LAG3 were equally distributed across effector and exhausted T cell populations. Interestingly, we were able to capture patient-specific heterogeneity of gene expression in T cells, suggesting the possibility of individualized T cell gene signatures present in PDAC tumors. We used mass cytometry and immunostaining to validate our transcript-based findings.

Conclusion: Overall, we have successfully performed robust in-depth profiling using single-cell sequencing of PDAC tumors from fine needle biopsies. TIGIT, but not other immune checkpoints, correlates with T cells exhaustion in tumors, revealing an important biological function of this relatively understudied checkpoint. Analysis of our results identified patient-specific heterogeneity of key signaling pathways in different cell compartments of PDAC tumors that have to potential to be leveraged for precision medicine.

Citation Format: Nina Steele, Eileen Carpenter, Samantha Kemp, Veerin Sirihorachai, Stephanie The, Lawrence Delrosario, Jenny Lazarus, El-ad Amir, Valerie Gunchick, Carlos Espinoza, Samantha Bell, Lindsey Harris, Valerie Irizarry-Negron, Dan Paglia, Justin Macchia, Fatima Lima, Angel Ka Yan Chu, Heather Schofield, Erik Jan Wamsteker, Richard Kwon, Allison Schulman, Anoop Prabhu, Ryan law, Arjun Sondhi, Katelyn Donahue, Hari Nathan, Clifford Cho, Michelle Anderson, Vaibhav Sahai, Costas Lyssiotis, Benjamin Allen, Arvind Rao, Weiping Zou, Filip Bednar, Timothy Frankel, Marina Pasca di Magliano. Multimodal mapping of the immune landscape in human pancreatic cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3442.