10.04 Spatial transcriptomics identifies metabolic dysregulation as a key driver of T cell exclusion in esophageal adenocarcinoma

• Published in: Journal for immunotherapy of cancer 2024-03, Vol.12 (Suppl 1), p.A5-A6
• Main Authors: Sanders, J, Bos, EN, Hahn, N, Nijman, I, van Laarhoven, HWM, van der Peet, DL, Van den Bossche, J, de Gruijl, TD, Derks, S
• Format: Article
• Language: English
• Subjects: Antigen presentation; Cancer; Esophageal cancer; Esophagus; Lymphocytes; Metabolism; Oral Presentations; Tumors
• ISSN: 2051-1426
• DOI: 10.1136/jitc-2024-ITOC10.9

BackgroundSuccess of neoadjuvant chemoradiotherapy (nCRT) in esophageal adenocarcinoma (EAC) is dependent on the level of activation of the tumor immune microenvironment (TIME). Patients with a complete pathological response ((pCR), 20% of EACs) have higher intratumoral CD8 T cell levels and a higher CD8:CD163 ratio compared to patient with a non-pCR. To improve response to nCRT we have to identify and target the mechanism EACs use to keep CD8 T cells out.Materials and MethodsTo this end we used tissues of patients with high vs. low intratumoral CD8 T cell density (multiplex immunohistochemistry) from a previous study for spatial whole transcriptomics (Nanostring GeoMx DSP) to characterize the transcriptome of cancer cells in CD8 high vs. low tumor areas. Regions of interest (ROI) were determined based on T cell density (CD3+) and analyzed for whole transcriptome of both tumor cells (PanCK+) and adjacent immune cells (CD45+) separately. Transcriptional differences were validated using bulk transcriptome data from publicly available data bases (TCGA) and Single Cell ENergetIc metabolism by profiling Translation inhibition (SCENITH) using fresh resection material (n=4).ResultsWhole transcriptome analyses of CK+ ROIs (cancer cells) identified that they clustered separately based on T cell infiltration status, indicating that CK+ cells in inflamed EACs are transcriptionally distinct from those in non-inflamed EACs. Differential gene analysis showed that antigen presentation pathway genes (HLA-A, HLA-B, HLA-C, and B2M) were upregulated in CD8-high EACs, whereas tumors with CD8-low EACs overexpressed genes associated with attraction and M2-like polarization of macrophages, such as IDO1, CXCL5, and CSF2. Differential pathway analyses indicated that the largest difference between CD8-high and CD8-low EACs was related to lipid metabolism. CD8 high tumors show high activity of immune pathways, such as antigen presentation and IL7 signaling, whereas the CD8 low tumors were characterized by high activity of the HDL and chylomicron remodeling pathways. Pathway analysis of the CD45+ compartment identified upregulation of the PGC1a pathway in CD8 T cell low tumors, a known driver of mitochondrial biogenesis and associated with suppressive myeloid cells. The correlation between CD8 T cell status and mitochondrial metabolism was confirmed by publicly available transcriptional data from The Cancer Genome Atlas (TCGA) showing high OXPHOS and fatty acid oxidation to be associ