Abstract 2773: Deep profiling of the PD-1/PD-L1 pathway in 10000 cancers revealed changes in the immune cell composition between cancer entities

Abstract

Abstract Immune checkpoint inhibitor (CPI) therapies targeting the PD-1/PD-L1 pathway have shown remarkable results in a rising number of different tumor entities. The likelihood of a positive response rate to CPIs depends - amongst other factors - on the PD-L1 expression level in tumor cells and tumor infiltrating immune cells. PD-L1 expression on different cell types is known to impact both the function and the composition of immune cells in the tumor microenvironment. However, this process is not fully understood. To study the impact of PD-L1 expression on the intratumoral immune environment, a multiplex fluorescence immunohistochemistry approach was used enabling the simultaneous analysis of 15 different antibodies on a set of tissue microarrays containing samples from more than 10000 neoplasms from more than 100 different tumor types and subtypes. Image analysis was performed using a U-net deep learning algorithm for cell detection and the “R” software package for statistical analysis. PD-L1 expression was measured on tumor cells (AE1/AE3+), subsets of macrophages (CD68+CD163+/CD68+iNOS+) and dendritic cells (CD11c+). To address the complexity of the tumor microenvironment, PD-1 expression was measured on B-cells (CD20+) and T-cell subsets, including cytotoxic T-cells (CD3+CD8+), T-helper cells (CD3+CD4+), regulatory T-cells (CD3+CD4+FOXP3+) and their memory (CD45RO+), activated (CD7-) or expanding (ki67+) phenotypes. Although PD-L1 expression varied significantly between cell types, patients and different tumor entities, the number of PD-1+ T-cells was consistently linked to PD-L1 expression. Combined analysis of cell densities, expression patterns, intensity measurements, interaction and distance analysis between immune cells and tumor cells revealed distinct changes in the immune cell infiltration pattern linked to PD-L1 expression. Previously uncharacterized immune cell-composition dynamics in clustered tumor phenotypes, according to the PD-L1 expression, were detected. This included significant associations between PD-L1+ tumor cells and PD-L1+ antigen-presenting cells and various PD-1+ T-cell subsets. In conclusion, deep profiling of 15 biomarkers in 10 000 cancers revealed complex differences in the composition of tumor infiltrating immune cells depending on the PD-L1 expression level of tumor cells and the mononuclear phagocyte system. Citation Format: Niclas Christian Blessin, Cheng Yang, Jonas Raedler, Julia Ebner, Tim Mandelkow, Ronald Simon, Christoph Fraune, Maximilian Lennartz, Andreas M. Luebke, Sarah Minner, Eike Burandt, Doris Höflmayer, Guido Sauter, Katharina Möller, Sören A Weidemann. Deep profiling of the PD-1/PD-L1 pathway in 10000 cancers revealed changes in the immune cell composition between cancer entities [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2773.