Abstract
INTRODUCTION: Immune checkpoint inhibition (ICI) has proven to be a major breakthrough in hepatobiliary cancers treatment. However, immune-related adverse events (irAEs) remain a major concern. The gut microbiome has been implicated in ICI efficacy; however, specific alterations in the multi-kingdom gut microbiota associated with severe irAEs are not well understood. OBJECTIVES: We aimed to identify the signatures of gut microbiota, fungi, and metabolites in patients with advanced hepatobiliary cancers with severe irAEs compared to those in patients experiencing mild or no irAEs. METHODS: We enrolled 168 patients with advanced hepatobiliary cancers between June 2018 and June 2022 (72 in the train set, 31 in test set 1, and 65 in test set 2). Multi-kingdom microbiota profiles were investigated using metagenomic, ITS2, and metabolomic datasets. RESULTS: The presence of severe irAEs was associated with significantly longer overall survival compared with the irAE-Mild and irAE-No groups. Patients with severe irAEs showed significant differences in the composition of bacteria and metabolites, but relatively few differences in fungi, and had more complex network associations of multi-kingdom gut microbiota compared with the irAE-Mild and irAE-No groups. A predictive model based on four bacteria and six metabolites simultaneously discriminated irAE-Severe from irAE-Mild and irAE-No with high accuracy. CONCLUSION: Patients with severe irAEs exhibited unique changes in microbiota-fungi-metabolite interactions. Gut microbiota- and/or metabolite-based algorithms could be used as additional tools for predicting severe irAEs and as potential prognostic markers in advanced hepatobiliary cancers.