Unraveling the heterogeneity of intercellular crosstalk in intrahepatic and extrahepatic cholangiocarcinoma using single-cell and spatial transcriptome profiling.

Abstract

50 Background: Extrahepatic cholangiocarcinoma (ECC) and intrahepatic cholangiocarcinoma (ICC) are two types of malignant tumors that arise from the bile duct epithelium at different locations. Currently, effective treatment options for these tumors are limited, and their microenvironmental differences remain poorly understood. Methods: We collected retrospective clinicopathological data of ECC (N=64) and ICC (N=153) patients from our hospital and integrated public database information to analyze the clinical characteristics, survival, and prognosis of ECC and ICC. We performed single-cell sequencing (scRNA-seq) and spatial transcriptomic techniques on three ECC and three ICC tumor tissues. Verification of scRNA-seq results was done using immunohistochemistry, flow cytometry, and polychromatic immunofluorescence. Additionally, we analyzed RNA and ATAC sequencing data for exhausted CD8+ T cells TEX obtained from ICC and ECC tissue samples. We used InferCNV to infer copy number variations (CNVs) of epithelial cells from ICC and ECC. Results: Our results showed that ECC and ICC differ in their clinical characteristics such as age, gender, bilirubin, CEA, and outcome. ECC had higher copy number amplification variation and exhibited inflammatory signaling and ongoing immunoediting, reflected in higher immune checkpoint expression such as PDCD1, CTLA4, and LAG3, and tumor infiltration with highly differentiated dysfunctional PDCD1+ LAG3+ CXCL13+ exhausted CD8+ T cells (TEX) and precursors of exhausted T (TPEX). RNA velocity analysis revealed different developmental trajectories of T cells in ECC and ICC, predominantly in naive/stem-like and memory T cells. We also identified a trefoil factor 3 (TFF3)-positive epithelial subpopulation in ECC. TFF3+ epithelial cells had closer interaction with CTHRC1+ fibroblasts, which induced more CD8+ T cells to TEX and TPEX. TFF3+ cells exhibited higher CNV scores and higher PD-L1 and PD-L2 expression, and co-localized with TEX and TPEX, providing one explanation for the immunosuppressive environment in ECC and suggesting a potential response to immune checkpoint therapy (ICT). Conclusions: Our study provides a comprehensive characterization of the cellular heterogeneity landscape of ECC and ICC, which can guide the optimization of personalized cancer immunotherapy. We identified potential biomarkers and targets for immunotherapies, and our findings shed light on the underlying mechanisms of tumor immune evasion and provide a foundation for further research in this area.