Background: The overall efficacy of immunotherapy for advanced biliary tract cancers (BTCs) was still poor, exploring the potential biomarkers associated with tumor response in advanced BTCs may further improve the clinical benefits from immunotherapy. Methods: Twelve patients from a phase 2 clinical trial (ChiCTR2000036652), which evaluated the efficacy and safety of sintilimab (anti-PD-1 antibody), plus gemcitabine and cisplatin as first-line treatment in advanced BTC patients, were included in the current biomarkers analysis. Exploratory points mainly focused on differentiated immune-related gene expression and immune pathway signatures using a 289-gene immune-related RNA panel sequencing on baseline tumor samples. Differential gene expression analysis was performed between responders (tumor regression) and non-responders (tumor extension) using Wilcoxon rank-sum test. Signature scores were calculated using the Gene Set Variation Analysis package with publicly available gene signatures. Results: Twelve patients (8 responders vs. 4 non-responders) who had available clinical response evaluations with RNA-sequencing data on baseline tumor samples were included in the current biomarkers analysis. Differential gene expression analysis revealed a significantly increased CXCL10, CXCL13 expression (p < 0.001) and decreased MAGEA12 expression (p < 0.001) in responder tumors compared with non-responder tumors. Differentiated genes were significantly enriched in cytokine-cytokine receptor interaction and interleukin-17 signaling pathway in responder tumors. Extensive analyses on immune pathway signatures showed that higher expression of IFN-γ-related genes (p = 0.048) and T cell-inflamed genes (p = 0.016) were associated with response to immunocombination therapy. Meanwhile, immune cell types score was also calculated, no notable differences were found between defined groups. Conclusions: Some differentially expressed genes were investigated in responder tumors compared with non-responder tumors. Higher expression of IFN-γ-related genes and T cell-inflamed genes could potentially predict the effect of immunocombination therapy. Potential biomarkers are needed to identify patients who respond to sintilimab plus gemcitabine and cisplatin in a larger validation cohort. Clinical trial information: ChiCTR2000036652.