Abstract

Background Emerging evidence links oral bacterial microbiota dysbiosis to cholangiocarcinoma (CCA). However, the oral mycobiome in CCA patients remains unexplored. This study aimed to characterize the oral mycobiome in CCA. Methods Saliva samples were collected from CCA patients and healthy controls (HCs). ITS rRNA sequencing was performed to analyze microbial diversity, community composition, and function prediction. Spearman correlation analysis was conducted to explore the associations between fungi and bacteria, and key clinical indicators. Results A total of 132 HC and 72 CCA saliva samples were included. Compared with the HCs, the CCA group exhibited reduced oral mycobiome diversity. The CCA group was enriched with nine genera, including Candida , Komagataella and Ascomycota_gen_Incertae_sedi , while the HC group was enriched with 36 genera, such as Fungi_gen_Incertae_sedis , Bipolaris and Cryptococcus . Correlation analysis revealed significant alterations in the oral bacteria-fungi interaction network in CCA. The diagnostic model based on five ASVs demonstrated high accuracy, with AUC values of 99.91% in the discovery phase and 98.33% in the validation phase. Conclusions This study is first to reveal the characteristics of the oral mycobiome in CCA, identify its closely association with oral bacteria, and suggest the potential of the oral fungi for non-invasive diagnosis of CCA.