Abstract C025: Identification of cell-cell signaling programs across the tumor-tumor microenvironment ecosystem and associated with clinical status in neoadjuvant osimertinib treated patient samples revealed by spatial profiling

Abstract

Listen

Translate article

Abstract Lung cancer is the leading cause of cancer mortality, and despite improvements in treatment, tumors typically respond incompletely and resume growth after acquisition of drug resistance. Recent completion of a neoadjuvant osimertinib Phase II trial treating patients with surgically resectable stage I-IIIA EGFR-mutated non-small cell lung cancer (EGFRm NSCLC) (NCT03433469) (Blakely et. Al, JCO 2024) has highlighted the importance of further identifying non-genomic mechanisms of persistence and resistance to targeted therapy. Here we analyzed these patient samples to identify transcriptionally regulated signaling patterns enriched at Residual Disease (RD) after neoadjuvant osimertinib treatment compared to treatment naïve samples (TN), as well as signaling patterns enriched at RD in patients who ultimately develop disease recurrence after surgical resection compared to those who remain disease-free. Spatially resolved transcriptomic sequencing was performed, using the 10X Genomics Visium platform, on 36 tissue sections (n=9 TN, n=19 RD, n=4 Progressive Disease (PD), n=4 Tumor Adjacent Normal), from 25 patients. TN and PD samples were from standard-of-care surgical resections. 18 of 19 RD samples were from NCT03433469. The Visium array spots that tile the sequencing capture area are 55 µm in diameter, and thus can contain an estimated 1-10 cells within each array spot. After quality control, 91,582 array spots remained for downstream analysis. We identified high confidence tumor array spots annotated both as “Cancer” by a board-certified pathologist and with tumor characteristic copy number variations. Through dimensionality reduction clustering, we identified 10 major array spot clusters, including the high confidence tumor array spot cluster. These clusters were further subset into 44 high resolution clusters of array spot niches, 32 of which were shared across at least nine patients. Because array spots are not single-cell resolution, these niches represent spots with similar mixed cell type composition and transcriptional signatures. We identified significant increases in the CCL14+ Endothelial Cell Niche, CXCL2+ Endothelial Cell Niche, Lymphocytes + Endothelial Niche, and CXCL12+ Fibroblast Niche in RD samples compared to TN. We identified a cell signaling program defined by spatially co-expressed ligand-receptor pairs associated with focal adhesion, enriched in PD samples. This same pattern was enriched in RD samples from patients who had recurrent disease after surgical resection compared to RD samples from patients who remained disease free. These data describe an increase in immune trafficking in RD samples, as well as identify an EGFR- alternative signaling pathway correlating with osimertinib resistance. Identification of spatially resolved, differential architectural organization and signaling patterns in RD and PD states will expand our understanding of targeted therapy resistance in EGFRm NSCLC, with the potential to identify additional clinically actionable targets. Citation Format: Whitney Tamaki, Daniel L. Kerr, Wei Wu, Grant Eilers, Anatoly Urisman, Yu-Ting Chou, Philippe Gui, Shigeki Nanjo, Johannes R. Kratz, David M. Jablons, Trever G. Bivona, Collin M. Blakely. Identification of cell-cell signaling programs across the tumor-tumor microenvironment ecosystem and associated with clinical status in neoadjuvant osimertinib treated patient samples revealed by spatial profiling [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Tumor-body Interactions: The Roles of Micro- and Macroenvironment in Cancer; 2024 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(22_Suppl):Abstract nr C025.