Abstract B053: RAS(ON) inhibition, but not chemotherapy, sensitizes RAS mutant PDAC to immunotherapy via tumor microenvironment remodeling in preclinical models

Abstract

Pancreatic ductal adenocarcinoma (PDAC) remains largely refractory to immune checkpoint blockade and combining anti-PD-(L)1 therapy with standard-of-care chemotherapy has not improved clinical outcomes over chemotherapy alone. Because oncogenic RAS drives both tumor growth and tumor-intrinsic immune suppression, we hypothesized that direct RAS inhibition would remodel the tumor microenvironment (TME) and sensitize PDAC tumors to immunotherapy more effectively than cytotoxic chemotherapy. Here, we evaluated the effects of gemcitabine plus nab-paclitaxel (GnP) and RAS(ON) inhibition on the TME and response to anti-PD-1 therapy in immuno-refractory, GEMM-derived KPCY models of KRAS G12D-mutant PDAC. In immuno-refractory, T cell-infiltrated tumors, anti-PD-1 monotherapy had no activity and GnP achieved tumor growth inhibition, whereas daraxonrasib, an investigational RAS(ON) multi-selective tri-complex inhibitor, induced transient tumor regressions. Flow cytometric analyses showed that daraxonrasib, but not GnP, significantly increased T cell infiltration and reshaped intratumoral immune cell composition. These findings suggest that RAS(ON) inhibition induces immune-contexture changes not recapitulated by GnP chemotherapy. Consistent with these immune changes, daraxonrasib plus anti-PD-1 induced complete regressions in 50% of mice, whereas GnP failed to sensitize tumors to anti-PD-1. Moreover, although adding GnP to daraxonrasib increased the durability of regressions compared to daraxonrasib alone, the triple combination of daraxonrasib, GnP, and anti-PD-1 did not confer further benefit over daraxonrasib plus anti-PD-1, indicating that chemotherapy did not enhance daraxonrasib-mediated immune sensitization. We next asked whether deepening RAS pathway inhibition, relative to that achieved with daraxonrasib monotherapy, could further enhance immune sensitization in PDAC tumors. The combination of daraxonrasib with zoldonrasib, an investigational RAS(ON) G12D-selective, covalent, tri-complex inhibitor, and anti-PD-1 achieved complete regressions in 90% of mice, supporting deeper RAS pathway inhibition as a strategy to enhance immunotherapy responsiveness. Finally, we extended our analyses to T cell-low KPCY models, in which daraxonrasib alone does not sensitize tumors to anti-PD-1 therapy. Combining RAS(ON) inhibition with GnP improved durability of response, as observed in T cell-infiltrated tumors, but did not sensitize to anti-PD-1. In contrast, the addition of anti-PD-1 to the daraxonrasib plus zoldonrasib RAS(ON) doublet enhanced the durability of response in the T cell-low models, as was seen in the T cell-infiltrated KPCY models. Together, these preclinical findings demonstrate that RAS(ON) inhibition, unlike a standard-of-care chemotherapy, remodels the PDAC TME and sensitizes immune-refractory tumors to anti-PD-1 therapy. These data support the clinical evaluation of RAS(ON) inhibitor-based immunotherapy combinations, including daraxonrasib and zoldonrasib, as potential, differentiated strategies to overcome immunotherapy resistance in PDAC.