Abstract

Abstract The establishment of immune heterogeneity in the tumor microenvironment (TME) is poorly understood, despite recent data that the success of immunotherapies is dictated by the immune environment of the tumor site. Pancreatic ductal adenocarcinoma (PDA) is characteristically devoid of CD8 T-cells and resistant to therapeutic intervention. However, a small subset of patients (15%) have tumors highly infiltrated by CD8 T-cells, correlating with improved overall survival. To better elucidate the determinants of immune heterogeneity in the PDA TME, we generated clones from spontaneous tumors harvested from KrasG12D+/-;Trp53R172H+/-;Pdx-1 Cre (KPC) mice, a genetically engineered mouse model of PDA. Using a panel of 17 tumor clones, we found the clones segregated in to two groups with differential immune cell infiltration upon implantation in congenic C57BL/6 mice. 7/17 tumor clones were categorized as “T-cell high,” with an immune infiltrate comprising CD8 T-cells and CD103+ dendritic cells (DCs). In contrast, the remaining 10 tumor clones were categorized as “T-cell low” lines, with the TME dominated by myeloid cells and macrophages, especially granulocytic myeloid-derived suppressor cells. Hypothesizing that increased T-cell infiltrate would render PDA sensitive to therapy, we treated two T-cell high and two T-cell low tumor clones with combination immunotherapy. Mice bearing T-cell high clones responded to therapy (7/7 and 4/7 mice cured) and formed protective memory responses against secondary tumor challenge, while none of the mice bearing T-cell low tumors responded to treatment (0/7 and 0/7 mice cured). At baseline, T-cell high tumors had similar proportions of functional CD8 T-cells as in T-cell low tumors. However, the proportion of activated CD44hiPD-1+ CD8 T-cells was significantly increased in T-cell high tumors (62.2% vs. 35.1% in T-cell low clones, p<0.0001), and predicted response to therapy across the panel of tumor clones. CD44hiPD-1+ CD8 T-cells trafficked in to the tumor site independently of CXCR3 (53.8% in control vs. 50.0% in anti-CXCR3 treated tumors, p=0.8), in contrast to the bulk CD8 T-cell population which required CXCR3 to enter the TME (7.5% in control vs 0.53% in anti-CXCR3 treated tumors, p<0.01). Furthermore, in Batf3 KO mice, the T-cell high TME lacked CD8 T-cells and resembled the T-cell low phenotype, including resistance to combined immunotherapy with 0/7 mice responding to treatment, indicating the requirement for CD103+ DCs in the TME for T-cell infiltration and tumor regression. Co-mixing experiments revealed that the T-cell low phenotype was dominant in the local – but not systemic – setting, suggesting a secreted protein was involved in establishing the T-cell low TME, and CXCL1 was identified as a top candidate after transcriptomic and epigenetic profiling (padj = 0.088, basemean = 772.79). Overexpression of CXCL1 in T-cell high tumors converted the TME to a T-cell low phenotype, while ablation of CXCL1 in T-cell low tumors converted the TME to a T-cell high phenotype and rendered T-cell low sensitive to combined immunotherapy (16/16 responders or 7/16 responders in two independent CXCL1 KO T-cell low clones). These data reveal tumor cell intrinsic factors as major determinants of immune heterogeneity in the TME and highlight the use of this panel of clonal tumor lines as a tool to probe TME heterogeneity to develop optimized, patient-specific therapies. Citation Format: Katelyn T. Byrne, Jinyang Li, Robert H. Vonderheide, Ben Stanger. Tumor cell intrinsic factors dictate immune cell infiltration and response to immunotherapy [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A054.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call