Correlations between Intestinal Microbiota and Clinical Characteristics in Colorectal Adenoma/Carcinoma

Abstract

Background: Most of colorectal cancer (CRC) cases are sporadic and develop along the adenoma-carcinoma sequence. Intestinal microbial dysbiosis is involved in the development of colorectal cancer. However, there are still no absolute markers predicting the progression from adenoma to carcinoma.

Aims: To investigate the characteristics of intestinal microbiota in colorectal adenoma and carcinoma patients and the correlations with clinical characteristics.

Methods: Fecal samples were collected from 154 colorectal carcinoma patients (CRC group), 20 colorectal adenoma patients (AD group), and 199 healthy controls (control group). The intestinal microbiota was investigated by 16S rRNA gene sequencing.

Results: Compared to the healthy controls, microbial diversity was dramatically decreased in AD/CRC. At the genus level, Acidaminococcus significantly decreased with the order of control-AD-CRC (P < 0.05). Parvimonas, Peptostreptococcus, Prevotella, Butyricimonas, Alistipes, and Odoribacter were the key genera in the network of colorectal adenoma/carcinoma-associated bacteria. Combination of the top 10 most important species, including Butyricimonas synergistica, Agrobacterium larrymoorei, Bacteroides plebeius, Lachnospiraceae bacterium feline oral taxon 001, Clostridium scindens, Prevotella heparinolytica, bacterium LD2013, Streptococcus mutans, Lachnospiraceae bacterium 19gly4, and Eubacterium hallii, showed the best performance in distinguishing AD patients from CRC (AUC = 85.54%, 95% CI: 78.83%-92.25%). The clinicopathologic features, including age, sex, tumor location, differentiation degree, and TNM stage, were identified to be closely linked to the intestinal microbiome in CRC.

Conclusion: Several intestinal bacteria changed along the adenoma-carcinoma sequence and might be the potential markers for the diagnosis and treatment of colorectal adenoma/carcinoma. Intestinal microbiota characteristics in CRC should account for the host factors.

Figure 1

Intestinal microbiome richness and diversity in control, AD, and CRC groups. (a) Venn diagram illustrating the total, unique, and shared number of OTUs predicted for AD and CRC group datasets. (b) Rarefaction curve of OTU. (c) Species accumulation boxplot. (d) Boxplot based on Good's coverage. (e) Alpha diversity analysis based on the Chao1 index in AD and CRC groups (P = 0.95). (f) Alpha diversity analysis based on the Shannon index in AD and CRC groups (P = 0.74). (g) Principal coordinate analysis (PCoA) based on the weighted UniFrac distance matrix in AD and CRC groups (Pr(>F) = 0.33). (h) Nonmetric multidimensional scaling (NMDS) based on the weighted UniFrac distance matrix in AD and CRC groups (stress = 0.202). (i) Beta diversity analysis by the weighted UniFrac distance in AD and CRC groups (P = 0.88). (j) Principal coordinate analysis (PCoA) based on the weighted UniFrac distance matrix among control, AD, and CRC groups. (k) Nonmetric multidimensional scaling (NMDS) based on the weighted UniFrac distance matrix among control, AD, and CRC groups (stress = 0.123).

Figure 2

Top ten dominant genera in control, AD, and CRC groups.

Figure 3

Specific bacterial taxa associated with colorectal adenoma/carcinoma. (a) Venn diagram illustrating the total, unique, and shared number of significant pairwise difference at the genus level among control, AD, and CRC groups. (b) The relative abundance of the significantly changed genera among the three groups. (c) The most differentially abundant taxa between AD and CRC patients by linear discriminant analysis (LDA) effect size (LEfSe) analysis. Red bars indicate taxa enriched in the CRC group. (d) Microbial association network for colorectal adenoma/carcinoma-associated genera. (e) Receiver operating curve (ROC) by random forest analysis for distinguishing AD patients from CRC patients. (f) Mean Decrease Accuracy (MDA) coefficients of the 10 most important species in distinguishing AD patients from CRC patients. (g) Mean Decrease Gini (MDG) coefficients of the 10 most important species in distinguishing AD patients from CRC patients.

Figure 4

Association between fecal microbiota and CRC clinical characteristics: (a) the correlation heatmap between intestinal microbiota and age of CRC patients; (b) the correlation heatmap between intestinal microbiota and sex of CRC patients; (c) the correlation heatmap between intestinal microbiota and the tumor location; (d) the correlation heatmap between intestinal microbiota and the tumor stage.