A multiple-dimension model for microbiota of patients with colorectal cancer from normal participants and other intestinal disorders

• Published in: Applied microbiology and biotechnology Vol. 106; no. 5-6; pp. 2161 - 2173
• Main Authors: Shen, Jian, Jin, Gulei, Zhang, Zhengliang, Zhang, Jun, Sun, Yan, Xie, Xiaoxiao, Ma, Tingting, Zhu, Yongze, Du, Yaoqiang, Niu, Yaofang, Shi, Xinwei
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.03.2022; Springer; Springer Nature B.V
• Subjects: Adenoma; Analysis; Applied Microbial and Cell Physiology; Biological markers; Biomarkers; Biomedical and Life Sciences; Biotechnology; Cancer; Colon; Colorectal cancer; Colorectal carcinoma; Diagnosis; Digestive system; Feces; Gastrointestinal diseases; Gastrointestinal tract; Genetic aspects; Intestinal microflora; Life Sciences; Microbial Genetics and Genomics; Microbiology; Microbiomes; Microbiota; Microbiota (Symbiotic organisms); Microorganisms; Prediction models; Relative abundance; Risk; Risk factors; Risk groups; Stability; China; Gut microbiota; Biomarkers; 16S rRNA; Large-scale sequencing; Colorectal cancer
• ISSN: 0175-7598; 1432-0614
• DOI: 10.1007/s00253-022-11846-w

Gut microbiota is a primary driver of inflammation in the colon and is linked to early colorectal cancer (CRC) development. Thus, a novel and noninvasive microbiome-based model could promote screening in patients at average risk for CRC. Nevertheless, the relevance and effectiveness of microbial biomarkers for noninvasive CRC screening remains unclear, and researchers lack the data to distinguish CRC-related gut microbiome biomarkers from those of other common gastrointestinal (GI) diseases. Microbiome-based classification distinguishes patients with CRC from normal participants and excludes other CRC-relevant diseases (e.g., GI bleed, adenoma, bowel diseases, and postoperative). The area under the receiver operator characteristic curve (AUC) was 92.2%. Known associations with oral pathogenic features, benefits-generated features, and functional features of CRC were confirmed using the model. Our optimised prediction model was established using large-scale experimental population-based data and other sequence-based faecal microbial community data. This model can be used to identify the high-risk groups and has the potential to become a novel screening method for CRC biomarkers because of its low false-positive rate (FPR) and good stability. Key points • A total of 5744 CRC and non-CRC large-scale faecal samples were sequenced, and a model was constructed for CRC discrimination on the basis of the relative abundance of taxonomic and functional features. • This model could identify high-risk groups and become a novel screening method for CRC biomarkers because of its low FPR and good stability. • The association relationship of oral pathogenic features, benefits-generated features, and functional features in CRC was confirmed by the study.