Abstract PO-013: Defining immune infiltrate heterogeneity and its role in ovarian cancer chemotherapy resistance using single cell RNA sequencing

Abstract

Abstract Objectives: Immune cell infiltration is correlated with therapy response in ovarian cancer (OC), but exceptions to this correlation are numerous, which could be due to heterogeneity in the immune cell composition and functional phenotype of immune cells. To better understand the immune landscape of OC and to develop biomarkers predictive of OC response to chemotherapy we initiated a prospective study to analyze the unique molecular and histopathological characteristics of tumor samples taken during primary debulking, interval debulking and at recurrence. Methods: We have enrolled over 100 women and have completed single cell RNA sequencing (scRNAseq), multiplex immunohistochemical assays, H&E scoring for tumor infiltrating lymphocytes (TILs) and NanoString molecular subtyping in 30 women. Platinum status in the cohort was: sensitive (20), resistant (4), refractory (3), and not determined (3). ScRNAseq was performed using the 10X genomics platform and multiple bioinformatic algorithms were applied to annotate cell types present in the sample. Comparisons between cell types identified by scRNAseq and multiplex IHC were performed and statistical correlations were made between clinical characteristics such as platinum resistance and molecular phenotypes. TIL scoring was performed using Salgado scoring criteria and PDL-1/PD-1 IHC staining was assessed based on Tumor proportion score (TPS) and Combined Proportion score (CPS). Results: ScRNAseq combined with multiplex IHC revealed extensive heterogeneity both within and between patients. PD-1 and PDL-1 genes were expressed in 23/30 (76%) patients across some cell types (% expression range: 1-22%) while 20/30 (66%) showed expression both in immune and epithelial cells. Highest expression of both genes was noted in 4/30 (12%) patients. PDL-1 gene levels by ScRNAseq demonstrated robust linearity across high and low expression ranges noted on IHC assays. ScRNAseq demonstrated an added advantage of being able to detect genes on tumor samples with absent PDL-1 IHC staining. Differential expression of PD-1/PDL-1 genes among 4 molecular subtypes showed highest expression level in immunoreactive group. Interestingly, 2 patients in this group did not express these genes indicating that molecular subtyping alone might not be predictive of immunotherapy response. Stromal TILs of 50-90% and 20-40% were observed in 4/30 (13%) and 10/30 patients (33%) respectively, although no correlation was noted between TIL scoring and level of PD-1/PDL-1 genes. Conclusions: ScRNAseq identifies more patients with PD-1/PD-L1 compared to IHC assays, suggesting single biomarker analysis might not be predictive of treatment response. Our study is ongoing and we will present our analysis of molecular phenotypes and correlations with clinical features such as platinum resistance. The goal of this project is to use comprehensive molecular analyses to identify patients most likely to benefit from immunotherapy and further understand the mechanism of immune evasion in OC. Citation Format: Shobhana Talukdar, Jason Cepela, Zenas Chang, Mihir Shetty, Ying Zhang, Marissa Macchietto, Christine Henzler, Sarah A. Munro, Christopher Tastad, Aaron Grad, Sally A. Mullany, Jun Woo, Jinhua Wang, Andrew Nelson, Boris A. Winterhoff, Timothy K. Starr. Defining immune infiltrate heterogeneity and its role in ovarian cancer chemotherapy resistance using single cell RNA sequencing [abstract]. In: Proceedings of the AACR Virtual Special Conference on Tumor Heterogeneity: From Single Cells to Clinical Impact; 2020 Sep 17-18. Philadelphia (PA): AACR; Cancer Res 2020;80(21 Suppl):Abstract nr PO-013.