Longitudinal integrated analysis of gut microbiota and metabolites in rectal cancer patients undergoing chemoradiotherapy
- 1Microbiota Laboratory, Clinical Division of Microbiota, Department of Gastroenterology and Hepatology, The First Medical Center of Chinese PLA General Hospital
- 2Chinese PLA Medical School
- 3Department of Radiation Oncology, The First Medical Center of Chinese PLA General Hospital
Abstract
Background: Longitudinal changes in the fecal microbiome and metabolome during concurrent chemoradiotherapy (CCRT) for rectal cancer remain incompletely characterized.
Methods: In this prospective single-center cohort, 19 patients with rectal cancer were enrolled and 17 were included in the longitudinal analysis after two prespecified sampling-window exclusions. Fecal specimens were scheduled at nine time points: R0, collected before the first dose of capecitabine and before radiotherapy, and R3 through R24 at three-fraction intervals during treatment. Twenty unmatched healthy volunteers provided one fecal specimen for a descriptive pretreatment comparison. Full-length 16S rRNA gene sequencing and untargeted metabolomics were performed. Longitudinal microbial counts were analyzed using negative-binomial generalized linear mixed models, and metabolite features were analyzed using linear mixed-effects models; participant was included as a random intercept, and P values were adjusted using the Benjamini–Hochberg false discovery rate (FDR).
Results: Nineteen genera and 27 of 176 analyzed species showed significant linear time trends during CCRT (FDR < 0.05). Parvimonas micra, Peptostreptococcus stomatis , and Gemella morbillorum decreased, whereas Lachnoclostridium pacaense, Blautia hydrogenotrophica, [Eubacterium] sulci , and Coprococcus eutactus increased. Of 2,873 analyzed metabolic features, 393 changed significantly over time. In a secondary same-cohort analysis, 17 of 50 correlation-selected taxon-metabolite pairs remained significant in time-adjusted mixed-effects models, including a positive association between C. eutactus and a feature annotated as stercobilin. Among 103 significant features with KEGG identifiers, three pathways met a nominal raw-P threshold, but none remained significant after FDR correction. Pretreatment patient samples had lower Shannon diversity and differed from healthy-control samples by ANOSIM; these comparisons were descriptive because the groups were not matched for age or sex. Five species shared the same case-control direction in this cohort and GMrepo reports, but only P. stomatis and G. morbillorum also showed significant longitudinal declines during CCRT.
Conclusion: This small, intensively sampled cohort identifies exploratory fecal microbial and metabolic trajectories during capecitabine-based CCRT. The findings do not establish causal mechanisms, microbial functions, disease specificity, or biomarker performance and require independent confirmation in larger cohorts.
Related
- Longitudinal Gut Microbiota Changes in Women with Breast Cancer-Related Lymphedema Undergoing a Lifestyle Intervention
- Longitudinal gut microbiome changes during the chemoradiotherapy component of total neoadjuvant therapy for rectal cancer
- Natural Compounds and Their Microbiota-Derived Metabolites as Modulators of the Microbiota–Gut–Brain Axis in Alzheimer Disease: Biochemical Mechanisms and Drug Discovery Insights—A Narrative Review
- TMAO at the Crossroads of the Gut–Brain Axis: Emerging Mechanisms Linking Microbiota-Derived Metabolites to Neuroinflammation and Cognitive Dysfunction
- Paeonidin- and Cyanidin-Rich Fractions Ameliorate Gut Microbiota and Serum Metabolites in Rats with Type 2 Diabetes
BibTeX
@article{wang2026longitudinal,
title = {{Longitudinal integrated analysis of gut microbiota and metabolites in rectal cancer patients undergoing chemoradiotherapy}},
author = {Lin Wang and Pei Zhang and Yujing Zhang and Yan Li and Baolin Qu and Lihua Peng},
journal = {Frontiers in Microbiology},
year = {2026},
month = oct,
doi = {10.3389/fmicb.2026.1934703},
url = {https://www.frontiersin.org/articles/10.3389/fmicb.2026.1934703/full},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Longitudinal integrated analysis of gut microbiota and metabolites in rectal cancer patients undergoing chemoradiotherapy
AU - Lin Wang
AU - Pei Zhang
AU - Yujing Zhang
AU - Yan Li
AU - Baolin Qu
AU - Lihua Peng
PY - 2026
DA - 2026/10/14
JO - Frontiers in Microbiology
DO - 10.3389/fmicb.2026.1934703
UR - https://www.frontiersin.org/articles/10.3389/fmicb.2026.1934703/full
AB - Background: Longitudinal changes in the fecal microbiome and metabolome during concurrent chemoradiotherapy (CCRT) for rectal cancer remain incompletely characterized. Methods: In this prospective single-center cohort, 19 patients with rectal cancer were enrolled and 17 were included in the longitudinal analysis after two prespecified sampling-window exclusions. Fecal specimens were scheduled at nine time points: R0, collected before the first dose of capecitabine and before radiotherapy, and R3 through R24 at three-fraction intervals during treatment. Twenty unmatched healthy volunteers provided one fecal specimen for a descriptive pretreatment comparison. Full-length 16S rRNA gene sequencing and untargeted metabolomics were performed. Longitudinal microbial counts were analyzed using negative-binomial generalized linear mixed models, and metabolite features were analyzed using linear mixed-effects models; participant was included as a random intercept, and P values were adjusted using the Benjamini–Hochberg false discovery rate (FDR). Results: Nineteen genera and 27 of 176 analyzed species showed significant linear time trends during CCRT (FDR < 0.05). Parvimonas micra, Peptostreptococcus stomatis , and Gemella morbillorum decreased, whereas Lachnoclostridium pacaense, Blautia hydrogenotrophica, [Eubacterium] sulci , and Coprococcus eutactus increased. Of 2,873 analyzed metabolic features, 393 changed significantly over time. In a secondary same-cohort analysis, 17 of 50 correlation-selected taxon-metabolite pairs remained significant in time-adjusted mixed-effects models, including a positive association between C. eutactus and a feature annotated as stercobilin. Among 103 significant features with KEGG identifiers, three pathways met a nominal raw-P threshold, but none remained significant after FDR correction. Pretreatment patient samples had lower Shannon diversity and differed from healthy-control samples by ANOSIM; these comparisons were descriptive because the groups were not matched for age or sex. Five species shared the same case-control direction in this cohort and GMrepo reports, but only P. stomatis and G. morbillorum also showed significant longitudinal declines during CCRT. Conclusion: This small, intensively sampled cohort identifies exploratory fecal microbial and metabolic trajectories during capecitabine-based CCRT. The findings do not establish causal mechanisms, microbial functions, disease specificity, or biomarker performance and require independent confirmation in larger cohorts.
SN - 1664-302X
ER -