Fecal microbiome and metabolome dynamics during immunotherapy-based total neoadjuvant therapy in rectal cancer: associations with treatment response and toxicity.

BACKGROUND: Immunotherapy-based total neoadjuvant therapy (iTNT) is a promising strategy for microsatellite-stable locally advanced rectal cancer (LARC), yet therapeutic response and treatment-related toxicity remain heterogeneous. Integrated fecal microbiome and metabolome profiling may provide non-invasive biomarkers and functional clues for optimizing iTNT.

METHODS: We conducted a longitudinal fecal multi-omics study using samples from patients with microsatellite-stable LARC enrolled in the TORCH trial (NCT04518280). A total of 102 fecal samples were collected before treatment, during treatment, and after completion of iTNT. Metagenomic sequencing and untargeted metabolomics were integrated to characterize longitudinal microbial and metabolic changes. We also examined baseline features associated with therapeutic response, and multi-omics signatures linked to hematologic and gastrointestinal toxicities. A murine tumor model treated with radiotherapy plus immunotherapy, with or without GABA supplementation, was used for functional testing of the response-associated metabolite.

RESULTS: iTNT induced longitudinal gut microbiome remodeling. This remodeling was characterized by altered community structure, increased alpha diversity, enhanced microbial network connectivity, enrichment of Firmicutes-associated taxa, and depletion of Bacteroidetes and Proteobacteria. Fecal metabolomic profiles also shifted during treatment, with prominent changes in amino acid-related pathways and significant concordance between microbial and metabolic profiles. Responders were enriched in several Firmicutes-associated genera, including Ruminococcus, Anaerostipes, and Coprobacillus. In contrast, non-responders showed enrichment of Klebsiella and response-associated metabolites including gamma-aminobutyric acid (GABA). Microbial functional and metabolomic pathway analyses showed convergent enrichment of arginine and proline metabolism, which includes an alternative GABA-related metabolic route. Functionally, GABA supplementation weakened the antitumor efficacy of radiotherapy plus immunotherapy and was accompanied by systemic T cell dysfunction. In addition, specific microbial taxa and fecal metabolic features were associated with hematologic toxicity and diarrhea severity, with baseline metabolites showing exploratory potential for toxicity stratification.

CONCLUSION: This study provides a longitudinal fecal microbiome-metabolome resource for iTNT in LARC and identifies candidate microbial and metabolic features associated with treatment response and toxicity. GABA was functionally supported as a response-associated immunomodulatory metabolite, while candidate microbial functional signals warrant further mechanistic validation.