Abstract Endometrial Cancer (EC) is the most common gynecological malignancy in the United States, with striking heterogeneity of risk factors, molecular identities, and outcomes. Two common subtypes of EC are Type 1, which display endometrioid histology and include those with MSI status, and Type 2, which arise from an atrophic background and are more severe in their morbidity. Despite significant research on checkpoint blockade therapies, such as PD-1/PD-L1 blockade to treat Type 1 ECs, variable response rates underscore the need for further immunomodulation. NK cells are cytotoxic innate lymphocytes that cooperate with other cells of the immune system. Despite years of investigation into basic NK cell biology, our understanding of the role(s) of NK cells in human solid tumors is still in its infancy. For example, in head and neck squamous cell carcinoma (HNSCC) indicate that a subset of NK cells has strong anti-tumor function, while a separate subset of NK cells lacks direct cellular cytotoxicity. Considering these findings, we used flow cytometry to evaluate NK cell phenotypes and frequencies in fresh EC patient tumor samples. Our data suggest a functional dichotomy of NK cell subsets with distinct phenotypes, present in both Type 1 and Type 2 EC tumors (Type 1: 35.5 +/- 10.65 % CD16+ vs 73.5 +/- 10.41 % CD16-, n=6, p=0.10; Type 2: 16.23+/- 3.61 CD16+ vs 83.77 +/- 3.65 CD16-, n=3, p<0.05). The first subset expresses CD16, which mediates antibody dependent cellular cytotoxicity, along with T-BET and EOMES, which are required for direct killing through Perforin and Granzymes and thus reminiscent of conventional NK (cNK) cells found in blood. The second subset is CD16- with decreased expression of T-BET and EOMES, and upregulated expression of tissue resident NK cell (trNK)-associated markers including CD200R1, CD103 and CD49a. To understand frequencies of these NK cell subsets in EC, we used immunohistochemistry for CD3, CD56, CD103, and Granzyme B on formalin-fixed, paraffin-embedded tissues from 10 patients: 4 Type 1, 4 Type 2, and 2 benign controls. We utilized CD56 to denote NK cells and found that NK cells infiltrated the Type 1 EC cases more compared to Type 2 (22.1 +/- 7.3% NK cells in Type 1 infiltrate vs 14.8 +/- 2% NK cells in Type 2 infiltrate, n=5, p=0.17). Further, NK cells expressing CD103, denoting trNK cells, could primarily be found in the glands of EC tumors, while NK cells expressing Granzyme B, denoting the cNK cells, were primarily in the stroma and largely absent in the glands (Granzyme B+: 78.6 +/- 3.7% Stroma vs 21.4 +/- 3.7% Glands, n=3, p<0.05; CD103+: 15.6 +/- 4.5% Stroma vs 84.3 +/- .64% Glands, n=3, p=0.076). This work suggests that an NK cell dichotomy exists within EC, with distinct phenotypes and localizations. Ongoing work will include functional studies to assess killing potential and cytokine production of the two subsets, along with single cell RNA sequencing to unravel which NK cells may be friend or foe in EC. Citation Format: Caprice D. Eisele, Courtney J. Riedinger, Jesse J. Kowalski, Bethany L. Mundy-Bosse, Paige M. Porrett, Patrick L. Collins, Casey M. Cosgrove, Aharon G. Freud. Uterine natural killer cells display a unique functional dichotomy in human endometrial cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1351.
Read full abstract