785 Estrogen-deprivation promotes Th1 polarization of tumor-associated T cells in a mouse model of high grade serous ovarian cancer

Background Immunotherapy has achieved long-term survival in patients with melanoma and other tumors, introducing a new paradigm in cancer treatment. Differential outcomes among men and women receiving immune checkpoint inhibitors implicate sex steroids as modulators of treatment response. Estrogen signaling has a profound impact on T cell function and has been shown to upregulate FoxP3 expression, promoting a suppressive regulatory phenotype. Conversely, estrogen deprivation promotes Th1 skewing, including increased IFN-γ production in response to antigen-specific stimulation. We hypothesize that immunomodulatory effects of estrogen deprivation will enhance immunotherapy outcomes. Our lab has previously demonstrated that IFN-γ levels in the TME predict response to immune checkpoint blockade (ICB) regimens in ovarian cancer models. CTLA4 but not PD1/PDL1 ICB combined with PARP inhibition (PARPi), an oral chemotherapeutic, significantly increased IFN-γ in the TME. Furthermore, IFN-γ was required for the durable survival benefit achieved with PARPi/anti-CTLA4. Here we test whether estrogen deprivation enhances IFN-γ production in the TME and response to PARPi/anti-PD1. Methods Five-week-old female FVB mice underwent oophorectomy, laparotomy without oophorectomy (sham), or no surgery (n = 5 per group). On day 10, mice were intraperitoneally challenged with 200,000 BR5-Akt syngeneic OC cells and randomly assigned to receive either PARPi/anti-CTLA4, PARPi/anti-PD1 or vehicle control treatment. PARPi (40mg/kg/day) was administered days 13-30 and 100 μg of anti-CTLA4 or 300 μg anti-PD1 was administered on D14. A second dose of anti-PD1 was given on D24. On day 30, peritoneal cells were analyzed by flow cytometry. Tumor burden was measured by IVIS. Results Oophorectomy was associated with a significant increase in IFN-γ production by tumor-associated CD4+ T cells [30.4% vs 8.2%, p = 0.016] and an increase in the proportion of central memory CD8+ T cells [59.3% vs 34.9%, p = 0.007] in response to PARPi/anti-PD1 compared with sham and no-surgery controls. In contrast, no differences in T cell phenotype or function was noted among groups receiving PARPi/anti-CTLA4. These changes were associated with a decrease in tumor burden in response to PARPi/anti-PD1 on D30. Conclusions Estrogen deprivation promotes Th1 polarization among tumor-associated T cells in response to PARPi/anti-PD1 treatment. With evidence that high levels of IFN-γ in the TME strongly correlate with survival, we predict that these effects will enhance treatment outcomes in response to PARPi/anti-PD1. This work presents a rationale for testing estrogen receptor modulators in combination with immune therapy agents and provides a potential mechanism to account for observed differences in patient outcomes.