Spatial multi-omics profiling of the primary lung microenvironment to elucidate the mechanism of osseous metastasis in advanced NSCLC through imaging mass cytometry.

e20546 Background: Non-small cell lung cancer (NSCLC) is one of the world's deadliest cancers, including lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) . Osseous metastasis (OM) is common metastatic event and represents one of the direst consequences of NSCLC, portending a grim prognosis. Therefore, it’s vital to explore and improve the understanding of the etiology and pathogenesis of OM in NSCLC. Anatomizing tumor microenvironment (TME) can provide new insights into the mechanisms of osseous metastases. And the TME distinctions of advanced NSCLC with or without OM at the single-cell protein level have not been fully explored. Methods: We collected 32 tissues biopsy samples and clinical characteristics from advanced NSCLC patients with or without OM. 32 tissues were divided in to 4 groups (n = 8, each group): LUAD-NC, LUAD-OM, LUSC-NC and LUSC-OM. Then, we used imaging mass cytometry (IMC) to explore the spatial proteomic features and the factors influencing the therapeutic effect. Results: A total of 193,246 cells with spatial information revealed by IMC depicted the differences in the TME. Compared to LUSC-NC, CAFs and LYVE1 + epithelial cells showed higher abundance in LUSC-OM, while CD8 + T cells, CX3CR1 + CD4 + T cells, CX3CR1 + M1 like macrophage and E-cad - Ki67 + proliferating epithelial cells showed lower abundance. Among all CAFs, mCAFs, vCAFs, tCAFs and LYVE1 + CAFs were higher infiltration in LUSC-OM. Compared to LUAD-NC, B cells, memory CD4 + T cells, TCF1 + CD4 + T cells, memory CD8 + T cells and EMT epithelial cells showed higher abundance in LUAD-OM, while CAFs showed lower abundance. In our study, 4 LUSC-OM , 4 LUSC-NC, 2 LUAD-OM and 1 LUAD-NC received the immunotherapy plus chemotherapy treatment. The efficacy assessment of LUSC-OM is 3 SD, 1 PD, while 4 PR in LUSC-NC and 3 PR in LUAD. The main reasons for treatment failure in LUSC-OM were the low abundance of PD-L1 + , HLA-DR + and E-cad + Ki67 + proliferating epithelial cells. In patients with effective treatment, the main positive factor influencing drug efficacy was the low abundance of CX3CR1 + CD8 + T cells. Moreover, the low abundance CD8 + TRM and Treg in preoperative TME were the high-risk factors for 1-year postoperative recurrence in LUSC-NC. Conclusions: This study determines the spatial multi-omics profiling of TME components at a single-cell resolution in NSCLC with OM. The TME in LUSC-OM presents an oligoimmune state, in which CAFs and LYVE1 + epithelial cells are high risk factors for OM. And the TME in LUAD-OM presents an immune-enrichment state with EMT epithelial cells paling important role in OM. Moreover, the complex TME affect the treatment efficacy and postoperative recurrence. In a word, the comprehensive analysis may contribute to the early identification of OM in NSCLC and the development of therapeutic options.