Abstract 1175: Spatially resolved transcriptomics to understand mechanisms of response to neoadjuvant chemotherapy

Abstract

Abstract Background: Clinical response to therapy of patients diagnosed with high-grade serous ovarian carcinoma (HGSC) is highly variable. Mechanisms underlying response to treatment are not well understood. We previously performed bulk multi-omics analysis aimed at identifying biological differences between poor (PR) and excellent responders (ER) to neoadjuvant chemotherapy (NACT). Although differences were found, these were not as extensive as we had anticipated, suggesting that important molecular differences between such tumors may not be fully evident by bulk analyses. Thus, we carried out in-depth spatial analysis of HGSC to explore heterogeneity of disease and determine how the dynamic behavior between the tumor and the tumor microenvironment varies between PR and ER groups. Methods: We selected a cohort of patients with highly annotated HGSC samples categorized by response to NACT (PR and ER) and analyzed pre-treatment tumor tissues using bulk RNA sequencing (bRNASeq) and Visium Spatial Gene Expression. The definition of PR was stable or progressive disease after 3-4 cycles and/or suboptimal interval cytoreduction after NACT. ER was complete response or only microscopic disease left at the time of interval surgery. Results: To generate expression profiles, we used an unsupervised approach and identified nine distinct clusters, with at least 20 cell types. A comparison of the cell populations identified more stroma-dominated cell groups in the PR group. ER tumors contained more immune-related areas, with a high proportion of T, B and natural killer cells. This finding supports the established role of cytotoxic immune cells in strengthening chemotherapy response and better overall survival. In comparison with bRNASeq, we observed a blending phenomenon that hides deeper heterogeneity of the tumor. Some pathways, such as epithelial-mesenchymal transition (EMT), were statistically enriched in specific clusters of PR compared with the ER group, proving that deep characterization of tumor heterogeneity can be achieved only when exploring specific clusters of cells. The spatial distribution of the clusters showed that in the PR group, the clusters tended to be physically larger and distributed throughout the whole tissue. The ligand-receptor analysis revealed that co-expression of specific ligand-receptor pairing exists depending on their geographical localization, with closer interactions leading to stronger co-expression patterns. Conclusion: Two important findings emerged: importance of the stromal component as a potential driver of poor response to NACT; and identification of differential clusters (distribution and composition) in ER compared with PR tissues. Comparison of the in situ technique with bRNASeq allowed characterization of specific cell subpopulations that might be important determinants of lack of response to therapy, representing candidate therapeutic targets. Citation Format: Elaine Stur, Sara Corvigno, Mingchu Xu, Ken Chen, Sanghoon Lee, Jinsong Liu, Emilly Ricco, Nicole Fleming, Emine Bayraktar, Daniel Kraushaar, Jianhua Zhang, Anil K. Sood. Spatially resolved transcriptomics to understand mechanisms of response to neoadjuvant chemotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1175.