The rationale of reprogramming TREM2+ macrophages for overcoming the resistance to immune checkpoint inhibitors in pancreatic adenocarcinoma.

e14649 Background: Pancreatic ductal adenocarcinoma (PDAC) resists to immune checkpoint inhibitors as single agents. TAMs expressing the triggering receptor expressed on myeloid cells (TREM2) protein were found to be a dominant macrophage population in PDAC; targeting TREM2 has been found to enhance the efficacy of immune checkpoint inhibitor-based therapies in preclinical studies of other malignancies. Here, we studied the role of targeting TREM2 in PDAC by testing the combination of non-depleting anti-TREM2 antibody and anti-PD-1 antibody in a mouse model of PDAC. We also examined the relationship between TREM2 expression and effector T cells in surgically resected human PDAC collected from a previous clinical trial treated with the neoadjuvant combination therapy of nivolumab (anti-PD-1 antibody), urelumab (anti-CD137 agonist antibody), and the pancreatic cancer GVAX vaccine. Methods: C57BL/6 mice were orthotopically implanted with KPC tumors. Anti-TREM2 was injected intraperitoneally (i.p.) once a week and anti-PD-1 was injected i.p. twice a week. Tumor volume was measured by ultrasound. TREM2 was stained on slides which were previously obtained from surgically resected PDACs following the neoadjuvant combination treatment of nivolumab, urelumab, and GVAX and stained for multiple immune cell markers with a multiplex staining-striping immunohistochemistry (IHC) technique. Results: Tumor growth post-tumor implantation was found to be significantly decreased by the anti-PD-1 and anti-TREM2 combination compared with anti-PD-1 alone, anti-TREM2 alone, or control treatment (p < 0.0001). Survival of tumor-bearing mice was significantly prolonged by the anti-PD-1 and anti-TREM2 combination when compared to anti-PD-1(p < 0.05), anti-TREM2(p < 0.05), and control (p < 0.01). Multiplex IHC analysis of human PDACs following neoadjuvant immunotherapy showed a higher density of granzyme B+CD8+ effector T cells correlated with higher density of TREM2+ cells (p = 0.0357). Conclusions: Our study demonstrates the anti-tumor efficacy of anti-TREM2 antibody in combination with an immune checkpoint inhibitor in mouse PDAC. However, the analysis of human PDAC following neoadjuvant immunotherapy suggests that TREM2 on TAMs may be activated following the activation of effector T cells. Taken together, our results suggest that TREM2+ TAMs should be reprogrammed instead of being depleted in the design of anti-TREM2-based therapy for PDAC.