Longitudinal gut microbiome dynamics in rectal cancer: associations with neoadjuvant therapy and tumor regression

Abstract Background The gut microbiome has emerged as a key modulator of colorectal cancer biology, influencing tumor progression and therapeutic response. However, most existing studies analyze colorectal cancer as a single entity despite biological and therapeutic differences between colon and rectal cancer. Consequently, longitudinal data linking microbiome dynamics to treatment response, particularly in rectal cancer, remain limited. Methods In this prospective cohort study, 31 patients with rectal cancer were included, of whom 17 received neoadjuvant therapy and 14 underwent primary surgery. A total of 110 stool and tissue samples were collected at defined time points before, during, and after treatment. Microbial composition and diversity were analyzed using 16S rRNA gene sequencing. Alpha and beta diversity metrics, differential abundance analyses, and longitudinal comparisons were performed to assess treatment-related microbiome alterations. Results Distinct microbial communities were observed between stool and tissue samples, with stool samples enriched in Bacteroidetes and tissue samples dominated by Firmicutes , Actinobacteria , and Verrucomicrobia . Neoadjuvant therapy was associated with a marked reduction in microbial diversity with no recovery before surgery, and community composition shifted significantly over the treatment course (patient-constrained PERMANOVA, p = 0.0016), with Clostridium XIVa becoming the dominant genus before surgery. Longitudinal analyses further revealed exploratory differences between patients with complete and incomplete tumor regression. Between diagnosis and surgery, patients achieving complete tumor regression (Dworak TRG 4) showed an increase in the propionate-producing genus Phascolarctobacterium , the only genus significant after correction for multiple testing, together with an exploratory rise in the butyrate-producing genus Roseburia . Higher Bacteroides in patients with incomplete regression was a descriptive observation that was not statistically significant ( p = 0.35, q = 0.71). No genus-level differences between response groups were detected at baseline after correction for multiple testing; however, the small sample size limits the ability to exclude baseline differences. Conclusions Rectal cancer treatment induces distinct and time-dependent alterations in the gut microbiome. Complete tumor regression was associated with an increase in short-chain fatty acid–producing taxa, led by the propionate producer Phascolarctobacterium , whereas higher Bacteroides in incomplete responders was a non-significant, exploratory observation. Given the growing role of organ-preserving strategies such as watch-and-wait, in which reliable identification of complete response is critical, these exploratory patterns warrant validation in larger cohorts.