Immunogenomic signatures to predict outcome in ovarian and endometrial cancers: Potential strategies in targeting the tumor immune microenvironment to improve response to immunotherapy.

Abstract

4 Background: Ovarian and MSS endometrial cancers are characterized by immunosuppressive microenvironment (TME) and low response to immunotherapy with checkpoint inhibitors (CPI). Targeting immunosuppressive factors within TMErepresents an attractive approach to enhance response to CPI. Therefore, we sought to investigate different immunogenomic signatures and immune cells within TME and correlate them with survival. Methods: We used whole transcriptome sequencing of matched tumor-normal samples from 38 uterine serous cancer and TCGA data of ovarian (n = 374) and endometrial cancers (n = 541). Immunogenomic signatures focusing on Transforming Growth Factor (TGFβ), 18-genes IFN-γ and myeloid signatures (CD47 and B7H4 expressions) and immune cell abundance were investigated. Gene expression score was calculated by averaging the normalized and log transformed individual gene read counts. The optimized score cut off was selected to best separate the survival in the pilot cohort. Then the score was tested in TCGA RNAseq datasets. Population abundance of tissue-infiltrating immune cells were estimated using MCPcounter R package from bulk transcriptome data. Results: Higher IFN-γ and lower TGF-β signatures predicted better survival for endometrial and ovarian cancers (p < 0.05). The impact of TGF-β was higher in MSI-H vs. MSS cancers (p = 0.013 vs. 0.09). High CD47 predicted poor survival in endometrial cancer. Combined IFN-γ and TGF-β signatures predicted survival in the ovarian and endometrial cohorts (p < 0.001). Combined IFN-γ and CD47 expression predicted survival in endometrial cancer (p = 0.033). Analysis of immune cell abundance revealed enrichment of monocytic lineage and neutrophils but paucity of cytotoxic T-cells, NK cells, dendritic cells and B-cells. Immune cell abundance is being correlated with survival. Conclusions: Our data support the role of immunogenomic markers in predicting survival. We are evaluating these markers in predicting response to CPI in a pilot cohort. Immunogenomic markers represent the tumor microenvironment, can potentially guide rationale combination immunotherapy.