Gut microbiota signatures associated with diagnosis and treatment in rectal cancer

Abstract Gut microbiota dysbiosis is recognized as a critical factor in colorectal cancer (CRC) progression; yet, its role in rectal cancer (RC) and post-treatment microbial status remains underexplored. We performed 16S rRNA sequencing on fecal samples from 77 RC patients, including initially diagnosed (iRC) and post-treatment (tRC) cases, alongside 83 healthy controls (HC). Microbiota composition, alpha and beta diversity, and predicted functional profiles were analyzed, and two separate random forest models were constructed to distinguish iRC and tRC patients from healthy controls, respectively. Microbial alpha diversity was significantly lower in iRC compared with HC, with distinct community structures observed in iRC and tRC. Key taxa, including Bacteroides , Fusobacterium , and Prevotella, were enriched in iRC, whereas beneficial butyrate-producing taxa such as Faecalibacterium and Subdoligranulum were depleted in iRC. The microbiota in tRC exhibited partial restructuring but did not revert to a healthy state. The random forest models achieved high discriminatory performance, with areas under the receiver operating characteristic curves (AUCs) exceeding 0.9. Notably, Prevotella abundance was positively correlated with serum levels of CEA, CA199, CA242, and CA50 in the iRC group, suggesting its potential as a microbial biomarker associated with disease status. This study identifies specific microbiota signatures associated with RC and post-treatment status, and suggests that combining microbial biomarkers with machine learning may provide a non-invasive approach for distinguishing microbial profiles across different clinical groups. Key points • Gut microbiota composition and diversity differ significantly among HC, iRC, and tRC groups. • Two pairwise random forest models showed high discriminatory performance for distinguishing iRC and tRC from HC. • Differential genera were associated with serum tumor markers, and predicted microbial functions differed among groups.