Microbial characteristics of bile in gallstone patients: a comprehensive analysis of 9,939 cases

Abstract

Introduction: The exact triggers of gallstone formation remain incompletely understood, but research indicates that microbial infection is a significant factor and can interfere with treatment. There is no consensus on the bile microbial culture profiles in previous studies, and determining the microbial profile could aid in targeted prevention and treatment. The primary aim of this study is to investigate the differences in microbial communities cultured from bile specimens of patients with gallstones.

Methods: We collected the clinical characteristics and bile microbial status of 9,939 gallstone patients. Statistical analysis was employed to assess the relationship between microbes and clinical features, and a random forest model was utilized to predict recurrence.

Results: Results showed a higher proportion of females among patients, with the age group of 60-74 years being the most prevalent. The most common type of gallstone was solitary gallbladder stones. A total of 76 microbes were cultured from 5,153 patients, with Escherichia coli, Klebsiella pneumoniae, and Enterococcus faecalis being the most frequently identified. Significant differences in microbial diversity and positive detection rates were observed across different age groups, types of gallstones, and recurrence status. Positive frequencies of E. coli, Enterococcus faecium, and K. pneumoniae varied significantly by age group and gallstone type. The microbial diversity in the recurrence group was significantly lower compared to the non-recurrence group. The recurrence rate was significantly higher in the group with single microbial species compared to those with no microbes or multiple microbes. For the recurrence group, there were significant differences in the frequencies of seven microbes (Aeromonas hydrophila, Enterococcus casseliflavus, Enterococcus faecium, E. coli, K. pneumoniae, Proteus mirabilis, Pseudomonas aeruginosa) before and after recurrence, with these microbes appearing in a higher number of patients after recurrence. Regression analysis identified patient age, stone size, diabetes, venous thrombosis, liver cirrhosis, malignancy, coronary heart disease, and the number of microbial species as important predictors of recurrence. A random forest model constructed using these variables demonstrated good performance and high predictive ability (ROC-AUC = 0.862).

Discussion: These findings highlight the significant role of microbial communities in gallstone formation and recurrence. Furthermore, the identified predictors of recurrence, including clinical factors and microbial diversity, may help develop personalized prevention and recurrence strategies for gallstone patients.

Keywords: bile; clinical characteristics; gallstone; microbes; recurrence.

Figure 1

Study design.

Figure 2

Sankey diagram of microbes in bile of all samples. We show the taxonomic ranks kingdom, phylum, class, order, family, genus.

Figure 3

Number of microbial species distribution in gallstone individuals (A) and microbial distribution of dominant species (B). Only the microbes with a frequency >100 individuals are displayed in (B).

Figure 4

Differences in frequency rate of various variables between microbial positive and negative groups. (A) age; (B) stone sizes; (C) types of gallstones and comorbidities. “*” indicates P < 0.05; “***” indicates P < 0.001. Only comparisons with significant statistical difference (P < 0.05) are shown.

Figure 5

Differences in microbial diversity (A) and detection (B) rates across various clinical characteristics. Only clinical characteristics with significant differences are shown in the figure, namely different age groups, types of gallstones, and recurrence status. Different superscript letters (a, b and c) indicate significant differences. *** indicates P < 0.001, ** indicates P < 0.01.

Figure 6

Differences in frequency rate of specific microbes among four different age groups (A) and different types of gallstones (B). Different lowercase superscript letters indicate significant differences among different groups (P < 0.05). *** indicates P < 0.001, ** indicates P < 0.01.

Figure 7

(A) Differences in recurrence frequencies among the groups with no, single, and multiple microbes' groups. Different lowercase superscript letters indicate significant differences among different groups (P < 0.05). *** indicates P < 0.001. (B) Venn diagram of the number of microbial species present before and after recurrence. (C) Distribution of the frequencies of microbes with significant differences before and after recurrence. The heatmap shows the distribution of frequencies for seven microbes under different conditions before (Pre) and after (Post) recurrence. The vertical axis lists the microbial species, while the horizontal axis represents different condition combinations. Post-0, Pre-0 indicates the number of patients who were negative for the microorganism both before and after recurrence; Post-0, Pre-1 indicates the number of patients who were positive before recurrence and negative after recurrence; Post-1, Pre-0 indicates the number of patients who were negative before recurrence and positive after recurrence; Post-1, Pre-1 indicates the number of patients who were positive both before and after recurrence. Numerical labels indicate specific frequency counts, and the shading of the color represents the magnitude of the counts, with darker colors indicating higher frequencies.

Figure 8

(A) Factors influencing recurrence. Multivariate binary logistic regression analysis was conducted to calculate odds ratios (ORs) and 95% confidence intervals (CI), assessing factors associated with recurrence. (B) Model variable importance based on MeanDecreaseGini and MeanDecreaseAccuracy. (C) ROC Curve for recurrence prediction model. ROC curve illustrating the performance of the model in terms of sensitivity and specificity.