<p>Pancreatic ductal adenocarcinoma (PDAC) is largely resistant to immune checkpoint blockade (ICB) due to its highly immunosuppressive tumor microenvironment. How to switch the “immune-cold” microenvironment to “immune hot” one is a clinical unmet need. We recently generated a genetically engineered PDAC mouse model (KAR) by combining K-Ras mutation and AT-rich interactive domain-containing protein 1A (ARID1A) inactivation in the pancreas. Compared to the KPC (K-Ras mutation + p53 inactivation) tumors, KAR tumors express higher C-X-C chemokine receptor 2 (CXCR2) and have more CXCR2⁺ neutrophils infiltration. Our previous study demonstrated that activation of CXCR2 by its cognate ligands promotes proliferation and drug resistance in PDAC cells. In addition, increase in tumor-associated neutrophils and high neutrophil-to-lymphocyte ratio are associated with worse prognosis in PDAC patients. Here, we developed a monoclonal antibody against human CXCR2 (huCXCR2-Ab) and tested its therapeutic effect in cells and in animals. Treatment with huCXCR2-Ab inhibited IL-8–induced downstream signaling, and the proliferation of human PDAC cancer cells. In addition, huCXCR2-Ab suppressed PDAC growth in an orthotopic nude mouse model. We next developed a monoclonal antibody against mouse CXCR2 (muCXCR2-Ab) and investigated its immune regulatory effect in vivo. We showed that muCXCR2-Ab reduced tumor burden, increased CD8⁺ T-cell infiltration, and decreased immunosuppressive neutrophils in the KAR mice. Moreover, muCXCR2-Ab augments the therapeutic efficacy of anti-PD-L1. Thus, dual targeting of CXCR2⁺ tumor cells and tumor-associated neutrophils remodels tumor microenvironment and enhances immunotherapy response in ARID1A-deficient PDAC, providing a new strategy for precision therapy of PDAC.</p>