Gut Bacterial Characteristics and Noninvasive Detection of Colorectal Lesions

Abstract

Background and objectives:The gut microbiota are frequently reported to be associated with colorectal cancer, while less attention has been paid to precancerous tumors. This study aimed to characterize the intestinal bacteria in patients with colorectal lesions and to assess the potential of bacteria as noninvasive biomarkers of colorectal tumors Methods: We prospectively collected and sequenced 463 fecal samples from Zhongshan Hospital, Xiamen University, by targeting 16S rRNA V3_V4 on a Hiseq instrument with PE250 reagents. We analyzed the gut bacterial communities, determined the bacterial characteristics, and constructed models to classify colorectal tumors after feature selection, especially for precancerous lesions.Results: There was a significant difference in fecal bacterial communities among the controls with normal colons (healthy subjects; HS) and the four stages of colorectal tumors. The fecal bacterial diversity increased in colorectal tumors. The phylum Firmicutes was significantly decreased, while Bacteroidetes was increased in colorectal tumors vs. HS. Correspondingly, a total of 81 genera, 589 operational taxonomic units, and 157 predicted pathways were remarkably different in relative abundances among the five groups. Relatively weak differences were observed among colorectal hyperplastic or inflammatory polyps (CRP), small adenomas (CRA), and advanced adenomas (Adv_CRA). Based on feature selection from genera, operational taxonomic units, pathways, and age, the models achieved an area under the receiver operating characteristic curve of 0.92 for classifying colorectal tumors vs. HS, 0.91 for the precancerous tumors vs. CRC, 0.80 for Adv_CRA vs. CRP, and 0.70 for CRA vs. CRP. Conclusions:Alterations in the bacterial diversity, composition, and predicted pathways were identified across multistep colorectal tumorigenesis. The selected bacterial features represent potential noninvasive predictors of colorectal tumors, especially in discriminating benign polyps and adenomas.