Hypoxia-activated prodrug and antiangiogenic therapies cooperatively treat pancreatic cancer but elicit immunosuppressive G-MDSC infiltration
Arthur Liu, Seth T. Gammon, Federica Pisaneschi, Akash Boda, Casey R. Ager, David Piwnica-Worms, David S. Hong, Michael A. Curran
Arthur Liu, Seth T. Gammon, Federica Pisaneschi, Akash Boda, Casey R. Ager, David Piwnica-Worms, David S. Hong, Michael A. Curran
View: Text | PDF
Research Article Angiogenesis Therapeutics

Hypoxia-activated prodrug and antiangiogenic therapies cooperatively treat pancreatic cancer but elicit immunosuppressive G-MDSC infiltration

Abstract

We previously showed that ablation of tumor hypoxia can sensitize tumors to immune checkpoint blockade (ICB). Here, we used a Kras+/G12D TP53+/R172H Pdx1-Cre–derived (KPC-derived) model of pancreatic adenocarcinoma to examine the tumor response and adaptive resistance mechanisms involved in response to 2 established methods of hypoxia-reducing therapy: the hypoxia-activated prodrug TH-302 and vascular endothelial growth factor receptor 2 (VEGFR-2) blockade. The combination of both modalities normalized tumor vasculature, increased DNA damage and cell death, and delayed tumor growth. In contrast with prior cancer models, the combination did not alleviate overall tissue hypoxia or sensitize these KPC tumors to ICB therapy despite qualitative improvements to the CD8+ T cell response. Bulk tumor RNA sequencing, flow cytometry, and adoptive myeloid cell transfer suggested that treated tumor cells increased their capacity to recruit granulocytic myeloid-derived suppressor cells (G-MDSCs) through CCL9 secretion. Blockade of the CCL9/CCR1 axis could limit G-MDSC migration, and depletion of Ly6G-positive cells could sensitize tumors to the combination of TH-302, anti–VEGFR-2, and ICB. Together, these data suggest that pancreatic tumors modulate G-MDSC migration as an adaptive response to vascular normalization and that these immunosuppressive myeloid cells act in a setting of persistent hypoxia to maintain adaptive immune resistance.

Authors

Arthur Liu, Seth T. Gammon, Federica Pisaneschi, Akash Boda, Casey R. Ager, David Piwnica-Worms, David S. Hong, Michael A. Curran

×

Figure 3

ICB therapy fails to improve therapeutic responses to combination TH-302 and αVEGFR-2.

Options: View larger image (or click on image) Download as PowerPoint
ICB therapy fails to improve therapeutic responses to combination TH-302...
Mice bearing 10-day established MT4-LA tumors were treated for 5 consecutive days with TH-302 and 2 doses of VEGFR-2 spaced 4 days apart. On day 15, tumors were resected and processed for analysis of tumor-infiltrating immune cells. (A) CD8+ T cells were analyzed for expression of Ki67 and TOX (n = 10–14 per group). Poststimulation, CD8+ T cells were analyzed for expression of interferon-γ (IFN-γ), tumor necrosis factor-α (TNF-α), PD-1, and LAG-3 (n = 4–5 tumors pooled per group, 3 independent experiments). Frequency and percentage of PD-L1–expressing (B) M-MDSCs and (C) G-MDSCs (n = 10–14 per group). (D) Survival of mice bearing orthotopic MT4-LA treated with TH-302, αVEGFR-2, αPD-1, and αCTLA-4. Untreated mice and those treated with combination TH-302 and αVEGFR-2 are first presented in Figure 2 (n = 10–14 per group). Lowercase Vs represent doses. (A–C) One-way ANOVA followed by Tukey’s correction for multiple comparisons. (D) Log-rank (Mantel-Cox) test. *P adj < 0.05, **P adj < 0.01, ***P adj < 0.001, ****P adj < 0.0001; data are mean ± SEM.