Immune checkpoint inhibitors (ICIs) have revolutionized cancer therapy, but response rates vary significantly (13-68%). Factors influencing ICI efficacy include tumor PD-L1 expression, tumor mutational burden, and interferon-gamma pathway activity. Growing evidence suggests the gut microbiota plays a role in modulating antitumor immunity and ICI response. Preclinical and clinical studies have shown associations between gut microbiota composition and ICI efficacy in various cancers (melanoma, renal cell carcinoma, non-small cell lung cancer). For example, higher microbial diversity has been linked to better response, while specific species like *Akkermansia muciniphila* and *Bifidobacterium* have shown promise in enhancing ICI efficacy. However, the highly variable nature of bacterial genomes, particularly microbial structural variations (SVs), complicates understanding these associations. SVs represent a layer of heterogeneity that can significantly influence bacterial phenotypes, even among closely related strains. Previous studies have shown links between microbial SVs and host metabolic levels, but their role in ICI response remains largely unexplored. This study aimed to systematically investigate the association between gut microbial SVs and ICI treatment outcomes.

A substantial body of research highlights the interplay between the gut microbiome and the efficacy of immune checkpoint inhibitors (ICIs). Studies have demonstrated that gut microbiota composition influences antitumor immunity and impacts ICI response in various cancers. Preclinical models using fecal microbiota transplantation (FMT) have shown that specific bacterial species can either enhance or hinder the effectiveness of ICIs. For instance, *Bifidobacterium* species have been shown to enhance CD8+ T cell priming, improving ICI response. Conversely, certain bacterial species have been linked to poor ICI response. Clinical studies have further supported these findings, showing a correlation between microbial diversity and ICI efficacy. Specific species like *Akkermansia muciniphila* and *Faecalibacterium prausnitzii* have been identified as potential predictive biomarkers for ICI response. However, these studies largely focused on the taxonomic composition of the microbiome. This study expands upon this knowledge by investigating the role of microbial structural variations (SVs), a more nuanced level of genomic heterogeneity, in predicting ICI treatment outcomes.

This study analyzed publicly available metagenomic sequencing data from seven independent cohorts encompassing a total of 996 ICI-treated patients. The cohorts included patients with melanoma, renal cell carcinoma (RCC), and non-small cell lung cancer (NSCLC) treated with various ICIs (anti-PD-1, anti-CTLA-4, or combinations). Raw sequencing data were preprocessed to remove low-quality reads and host DNA contamination. Microbial structural variations (SVs) were identified using SGV-Finder, a tool designed to detect both deletion (dSV) and variable (vSV) SVs from metagenomic data. The SVs were analyzed for associations with clinical outcomes including response (CR/PR vs SD/PD), progression-free survival at 12 months (PFS12), overall survival (OS), and immune-related adverse events (irAEs). Statistical analyses included logistic regression for binary outcomes (response, PFS12, irAEs), Cox regression for OS, and permutational multivariate analysis of variance (PERMANOVA) to assess associations between SV profiles and clinical outcomes. Meta-analyses were conducted to combine results across cohorts. To determine if correlations were independent from taxonomic abundances, associations at the level of species abundances were tested. Significant associations were further investigated to identify genes within the associated SV regions.

The study identified numerous significant associations between gut microbial SVs and ICI treatment outcomes. These associations were independent of bacterial species abundance, highlighting the importance of considering SVs as a separate factor. Specific SVs in multiple bacterial species, including *Akkermansia muciniphila*, *Roseburia formicigenerans*, and *Bacteroides caccae*, were associated with response, PFS12, and irAEs. Analysis of associated SV regions revealed enrichment of genes encoding enzymes involved in glucose metabolism. The most strongly associated species were *Akkermansia muciniphila* and *Roseburia* species. For melanoma, several SVs in *A. muciniphila* were associated with response and PFS12, including one containing genes encoding glycosyl transferases. In NSCLC and RCC, SVs in *Bacteroides* were linked to response and OS, with some containing genes associated with glycosyl hydrolases. These results suggest that specific genetic variations within gut bacteria may be important modulators of ICI efficacy and adverse events.

This study demonstrates that gut microbial structural variations (SVs) are significantly associated with immune checkpoint inhibitor (ICI) response in cancer patients. The findings are particularly notable because the associations observed were independent of the overall abundance of the bacterial species harboring the SVs. This highlights the importance of considering sub-genomic variation in microbiome research, as SVs can represent a crucial layer of functional heterogeneity. The enrichment of genes involved in glucose metabolism within the associated SV regions suggests potential mechanisms by which these SVs may influence ICI efficacy. Further research should focus on elucidating these mechanisms, potentially through metabolomic studies to identify relevant metabolic pathways. This research also indicates that SVs could serve as potential biomarkers for predicting ICI response and tailoring treatment strategies. The identification of specific SVs associated with positive or negative outcomes might aid in the development of microbiome-based therapies to improve ICI efficacy.

This study significantly advances our understanding of the gut microbiome's role in immune checkpoint inhibitor (ICI) therapy. The systematic analysis of microbial structural variations (SVs) revealed robust associations with ICI treatment outcomes, independent of species abundance. The identification of specific SVs and associated genes involved in glucose metabolism opens exciting avenues for future research, including the development of novel microbiome-targeted therapies to enhance ICI efficacy and mitigate adverse events. Larger, longitudinal studies are needed to validate these findings and explore the causal relationships between specific SVs and ICI response.

This study is retrospective and relies on publicly available datasets, which may introduce biases due to variations in study design, patient populations, and data collection methods. The relatively small sample sizes for certain cancer types and specific clinical outcomes may limit statistical power. The cross-sectional design makes it difficult to establish causal relationships, so prospective and interventional studies are necessary to confirm the findings. The lack of functional characterization of the identified SVs also limits the understanding of their specific roles in mediating ICI responses.