Multi-Kingdom Gut Microbiome Interaction Characteristics Predict Immune Checkpoint Inhibitor Efficacy Across...

An increasing number of studies have confirmed that the gut microbiota, especially bacteria, is closely related to the efficacy of immune checkpoint inhibitor (ICI) therapy. However, the effectiveness of multi-kingdom microbiota and their interactions in predicting the therapeutic effect of ICI therapy remains uncertain. We integrated extensive gut metagenomic databases, including 1712 samples of 10 cohorts from 7 countries worldwide, to conduct rigorous differential analysis and co-occurrence network analysis targeting multi-kingdom microbiota (bacteria, fungi, archaea, and virus). We ultimately identified two subtypes (C1 and C2) by employing a weighted similarity network fusion (WSNF) method. Subtype C2 exhibited higher microbial diversity, better treatment response, and improved prognosis compared to subtype C1. Notably, subtype C2 was associated with higher abundance of beneficial genera such as Bacteroides and Kluyveromyces, while subtype C1 contained potentially detrimental taxa like Malassezia. A multi-kingdom model incorporating 32 genera demonstrated superior predictive accuracy for ICI therapy efficacy compared to single-kingdom models. Co-occurrence network analysis revealed a more robust and interconnected microbiome in subtype C2, suggesting a stable gut environment correlates with effective ICI therapy efficacy. This study highlights the potential of a multi-kingdom signature in predicting the efficacy of ICI therapy, offering a novel perspective for personalized therapy in oncology.