Characterized by its dense desmoplasia, the PDAC tumor microenvironment (TME) is highly immunosuppressive. Largely orchestrated by stromal pancreatic stellate cells (PSCs), the desmoplastic reaction observed in PDAC results in tumor hypoxia, a key factor in disease aggressiveness and chemotherapy resistance (5). Moreover, sequestration of immune cells, including CD8+ T cells, B cells, and natural killer (NK) cells, limits effective immune surveillance (6–8). Consequently, recent trials have investigated stromal targeting agents as well as immune cell–based therapies for the treatment of PDAC. For example, in the phase Ib Stromal TARgeting for PAncreatic Cancer (STARPAC) trial, all-trans retinoic acid (ATRA) was used as a stromal targeting agent to render PSCs quiescent, rather than eliminate them, as well as modulate immune response as measured by pentraxin-3 (9, 10). ATRA in combination with gemcitabine-Nab-paclitaxel was found to be safe and tolerable in patients with advanced, unresectable PDAC, leading to a phase II trial, STARPAC2 (11). Clinical development of preclinically promising stromal-targeting and immune-modifying agents, such as pegvorhyaluronidase Aalfa, when combined with standard chemotherapy regimen (Nab-paclitaxel plus gemcitabine) has not lived up to its hope (12). Moreover, the spatial distribution of immune cells, such as NK cells, can be altered in response to stromal targeting in murine models of PDAC (6). Taken together, these results suggest that modulation of the TME and the subsequent cell-cell interactions in PDAC may have prognostic implications.

Cellular therapies, such as chimeric antigen receptor (CAR) T cells have demonstrated striking efficacy in hematological malignancies and thus have received much attention as a potential treatment for solid tumors, including PDAC, with targets such as CEACAM7 and Claudin18.2 (3, 13). However, substantial challenges persist, including the effective trafficking of CAR-T cells to sites of malignancy. Furthermore, the heterogeneous PDAC TME may result in cellular dysfunction and exhaustion upon arrival (14, 15). Notably, one of the largest challenges facing CAR-T cell therapy is graft-versus-host disease (GvHD) in which the transplanted cells react to off-target, normal patient tissue, leading to adverse events (16).

Since the development of autologous CAR-T cells is expensive and time consuming and further work is required to reduce potential GvHD invoked by allogenic CAR-T cells, attention has fallen on alternative cell types as the foundation for cell-based therapies. Thus, NK and natural killer T (NKT) cells with a unique opportunity to reduce GvHD, while exhibiting rapid inherent cytolytic capabilities, have been developed to enhance the safety and efficacy of CAR-targeted cell therapy (17, 18).

The distinct immune cell subset NKT cells are characterized by an invariant α-chain T cell receptor. Unlike T cells, NKT cells engage in target recognition through engagement of the CD1d receptor on antigen presenting cells (Figure 1) (19). This mechanism of action limits the risk of GvHD typically induced by major histocompatibility complex (MHC) class I involvement. NKT cells bridge the gap between the innate and adaptive immune systems, possessing several NK-like functions, such as rapid cytokine secretion and cellular cytotoxicity (17, 18). Additionally, these effector cells are capable of intricate cellular crosstalk between cells of both the innate and adaptive immune system, promoting activation of dendritic and NK cells, B and T cells, as well as macrophages and neutrophils (20). Taken together, these attributes make NKT cells a very attractive therapeutic option.

CAR-invariant NKT cells engage via the CD1d receptor on malignant and antigen-presenting cells. CAR-invariant NKT cells target malignant cells through engagement of their CAR with the target antigen present on the cancer cell surface. Binding induces rapid cytokine release, stimulating recruitment of additional immune cell subsets, such as T cells and NK cells. In addition, CAR-iNKT cells induce direct tumor cell cytolysis through perforin release. CAR-iNKT cells may also engage with malignant cells through invariant NKT TCR engagement with tumor antigens presented via the CD1d receptor. While some tumoral cells express CD1d, this receptor-ligand interaction is largely mediated by antigen-presenting cells. Signaling through CD1d does not involve engagement of the MHCs and thus prevents induction of GvHD, a common adverse event observed in response to allogeneic CAR-T cell treatment.