The pancreatic cancer immune tumor microenvironment is negatively remodeled by gemcitabine while TGF-β receptor plus dual checkpoint inhibition maintains antitumor immune cells

Abstract

Pancreatic ductal adenocarcinoma (PDA) tumors have a highly immunosuppressive desmoplastic tumor microenvironment (TME) where immune checkpoint inhibition (ICI) therapy has been exceptionally ineffective. Transforming growth factor-β (TGF-β) receptor activation leads to cancer and immune cell proliferation and phenotype, and cytokine production leading to tumor progression and worse overall survival in PDA patients. We hypothesized that TGF-β receptor inhibition may alter PDA progression and antitumor immunity in the TME. Here, we used a syngeneic preclinical murine model of PDA to explore the impact of TGF-β pathway inhibitor galunisertib (GAL), dual checkpoint immunotherapy (anti-PD-L1 and CTLA-4), the chemotherapy gemcitabine (GEM), and their combinations on antitumor immune responses. Blockade of TGF-β and ICI in immune-competent mice bearing orthotopically injected murine PDA cells significantly inhibited tumor growth and was accompanied by antitumor M1 macrophage infiltration. In contrast, GEM treatment resulted in increased PDA tumor growth, decreased antitumor M1 macrophages, and decreased cytotoxic CD8+ T cell subpopulation compared to control mice. Together, these findings demonstrate the ability of TGF-β inhibition with GAL to prime antitumor immunity in the TME and the curative potential of combining GAL with dual ICI. These preclinical results indicate that targeted inhibition of TGF-β may enhance the efficacy of dual immunotherapy in PDA. Optimal manipulation of the immune TME with non-ICI therapy may enhance therapeutic efficacy.

Keywords: CTLA-4; PD-L1; galunisertib; gemcitabine; immune cells; immunotherapy; pancreatic cancer; tumor microenvironment.

FIGURE 1

Schematic and experimental design for tumor generation and treatments. C57BL/6 mice were orthotopically injected with luciferase-expressing KPC tumor cells (KPC-luciferase cells) in Matrigel. After 3 days, mice were randomly divided into different groups (n = 10 per group) and treated with gemcitabine (GEM), galunisertib (GAL), combined murine anti-PD-L1/anti-CTLA-4 antibodies (ICI), and their combinations for 3 weeks to investigate the efficacy of treatments. ICI, immune checkpoint inhibition; IgG, immunoglobulin G; IP, intraperitoneal.

FIGURE 2

Monitoring tumor activity with bioluminescent imaging (BLI). (A) Weekly in vivo tumor activity and growth assessment was obtained by BLI, n = 8–10. (B) Quantification of BLI. Relative luminescence flux is represented as the normalization of luciferase activity of Week 3 to Week 1. (C) Number of regressed and progressed tumors on the Day 21. (D) Tumor weights at the endpoint. The data is displayed as mean ± SEM, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. GAL, galunisertib; GEM, gemcitabine; ICI, immune checkpoint inhibition.

FIGURE 3

Frequency of CD4+ and CD8+ T cells in pancreatic tumors. (A) CD3+ T cells absolute numbers. (B) Frequencies of CD3+ cells with respect to CD45+. (C) Frequencies of intratumoral CD4+ cells with respect to CD45+. (D) Frequencies of intratumoral CD8+ cells with respect to CD45+ and (E) their CD8+/CD4+ ratio. The data is displayed as mean ± SEM, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. GAL, galunisertib; GEM, gemcitabine; ICI, immune checkpoint inhibition.

FIGURE 4

Frequency of M1 and M2 cells in pancreatic tumors. (A) TAMs absolute numbers. (B) Frequencies of TAMs with respect to CD45+. (C) Frequencies of M1 cells with respect to CD45+. (D) Frequencies of M2 cells with respect to CD45+ and (E) their M1/M2 ratio. The data is displayed as mean ± SEM, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. GAL, galunisertib; GEM, gemcitabine; ICI, immune checkpoint inhibition; TAM, tumor-associated macrophage.