Genomic signature analysis and survival outcomes in cholangiocarcinoma (CCA) using oncology research information exchange network (ORIEN) database.

Abstract

4101 Background: Immunotherapy (IO) is standard of care for advanced CCA patients (pts) regardless of PD-L1 expression. Lack of biomarkers that predict prolonged benefit or resistance to therapies are unmet gaps in this field that we attempted to fill using transcriptional and survival data in the ORIEN dataset. Methods: Through an ORIEN intermember project, we accessed an avatar dataset for clinical characteristics and transcriptional data of the CCA cohort. We first identified clinical factors highly associated with transcriptional variability and conducted principal component analysis (PCA), focusing on factors linked to the first principal component (PC1). Both specimen site and type (p<0.01) were found highly associated with PC1 and were included as co-variates in subsequent models. Then, immune deconvolution using XCell was performed on the ORIEN-CCA RNA-sequencing cohort to determine the transcriptome portion attributed to effector T cell (Teff) and regulatory T cell (Treg) for each patient. Teff:Treg ratios were calculated and used to divide patients into cold (low Teff:Treg) and hot (high Teff:Treg) cohorts based on their median (mTeff:Treg = 0.0004). Differential gene expression analysis comparing hot and cold tumors yielded 72 significantly differentially expressed genes (FC <1.5, adjusted p<0.05). Gene Set Enrichment Analysis (GSEA) of transcriptomic changes revealed enrichment for immune and metabolic pathway alterations. Results: The ORIEN avatar dataset comprised of 78 pts, with 39 designated as immune hot and 39 as immune cold (demographics in Table). Median overall survival (mOS) was 3.17 years in hot and 2.65 years in cold group . These groups exhibited significantly different Teff:Treg ratio (p<0.0001). Differential gene expression identified 72 altered genes, including upregulated immune genes (e.g., IL2RA, CD80, GBP5) in the hot group and metabolic genes (e.g., AZGP1) in the cold group. GSEA of these genes revealed immune pathway enrichment in hot tumors, with increased antigen presentation, NK cell function, and B-cell activation. Surprisingly, cold tumors were enriched of pathways involved in lipogenesis, oxidation, and proliferation. Conclusions: ORIEN-CCA cohort analysis revealed immune hot and cold tumors with distinct transcriptional alterations and differences in mOS. Correlating these alterations with IO use, and further studies of enriched pathways in cold tumors are warranted. [Table: see text]