
==== Front
Open Forum Infect Dis
Open Forum Infect Dis
ofid
Open Forum Infectious Diseases
2328-8957
Oxford University Press US

10.1093/ofid/ofae506
ofae506
Infectious Diseases in Special Populations
Major Article
AcademicSubjects/MED00290
Gut Microbial and Metabolic Features Associated With Clostridioides difficile Infection Recurrence in Children
Li Xiaolu Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Xiao Fangfei Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Wang Xufei Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Ye Lin Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Xiao Yongmei Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Li Dan Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Zhang Ting Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
Gut Microbiota and Metabolic Research Center, Institute of Pediatric Infection, Immunity and Critical Care Medicine, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

Wang Yizhong Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children's Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
Gut Microbiota and Metabolic Research Center, Institute of Pediatric Infection, Immunity and Critical Care Medicine, School of Medicine, Shanghai Jiao Tong University, Shanghai, China

X. L. and F. X. contributed equally to this work.

Correspondence: Yizhong Wang, PhD, Department of Gastroenterology, Hepatology and Nutrition, Shanghai Children’s Hospital, School of Medicine, Shanghai Jiao Tong University, 355 Luding Road, Shanghai 200062 China (wangyz@shchildren.com.cn); Ting Zhang, MD, PhD, Department of Gastroenterology, Hepatology, and Nutrition, Shanghai Children’s Hospital, School of Medicine, Shanghai Jiao Tong University, 355 Luding Road, Shanghai 200062, China (zhangt@shchildren.com.cn).
Potential conflicts of interest. All authors: No potential conflicts of interest.

9 2024
03 9 2024
03 9 2024
11 9 ofae50606 4 2024
29 8 2024
02 9 2024
24 9 2024
© The Author(s) 2024. Published by Oxford University Press on behalf of Infectious Diseases Society of America.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs licence (https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any medium, provided the original work is not altered or transformed in any way, and that the work is properly cited. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.

Abstract

Background

Recurrent Clostridioides difficile infection (CDI) is a critical clinical issue due to the increase in incidence and difficulty in treatment. We aimed to identify gut microbial and metabolic features associated with disease recurrence in a group of pediatric CDI patients.

Methods

A total of 84 children with primary CDI were prospectively enrolled in the study. Fecal samples collected at the initial diagnosis were subjected to 16S rRNA gene sequencing and targeted metabolomics analysis to profile the bacterial composition and metabolome.

Results

Twenty-six of 84 (31.0%) pediatric CDI patients experienced recurrence. The alpha diversity of the fecal microbiota was significantly lower in the recurrent group than in the nonrecurrent group, and the beta diversity was different from that of the nonrecurrent group. Taxonomic proﬁles revealed that the relative abundances of multiple bacterial taxa significantly differed between the recurrent and nonrecurrent groups. Linear discriminant analysis effect size analysis identified several bacterial genera that discriminated between recurrent and nonrecurrent groups, including Parabacteroides, Coprococcus, Dialister, and Clostridium. Recurrent bacteria presented lower abundances of several short-chain fatty acid (SCFA)–producing bacteria (Faecalibacterium, Butyricicoccus, Clostridium, Roseburia, and Ruminococcus), which were correlated with reduced fecal SCFA levels. In addition, several bile acids, including lithocholic acid (LCA), 12-ketoLCA, trihydroxycholestanoic acid, and deoxycholic acid, were decreased in recurrent patients.

Conclusions

Our study suggests that the differing gut microbiota profiles in pediatric CDI patients may contribute to disease recurrence by modulating SCFA concentrations and bile acid profiles. The gut microbiota and metabolite signatures may be used to predict disease recurrence in children with CDI.

bile acids
gut microbiota
metabolites
recurrent Clostridioides difficile infection
short-chain fatty acids
National Natural Science Foundation of China 10.13039/501100001809 32370951 81900472 Natural Science Foundation of Shanghai Municipality 10.13039/100007219 22ZR1451800 Shanghai Municipal Health Commission 10.13039/100017950 20214Y0349 Shanghai Children's Hospital 2021YQ03 School of Digestive Science Shanghai Jiao Tong University School of Medicine 10.13039/501100008233 Xinyi Digestive Disease Research Foundation KY-2023-08-01
==== Body
pmc Clostridioides difficile is the most common opportunistic pathogen responsible for healthcare-associated infection in both adults and children [1]. The symptoms of C difficile infection (CDI) range from self-resolving diarrhea to life-threatening fulminant colitis and toxic megacolon [2]. The disease is attributed to the production of 2 major virulence factors, toxins A (TcdA/enterotoxin) and B (TcdB/cytotoxin), by toxigenic C difficile strains, which can bind to intestinal epithelia to induce cell death and severe inflammation [3]. Due to increasing morbidity and mortality, CDI is a serious substantial health and economic burden on the healthcare system worldwide [4–6].

The development of CDI is associated with increased susceptibility to C difficile caused by alterations in the gut microbiota composition and function [7, 8]. Previous studies have revealed that either asymptomatic colonization or infection with C difficile in the human gut is associated with a loss of gut microbiota diversity [7, 8]. In CDI patients, the relative abundances of taxa of the phyla Bacteroidetes and Firmicutes generally decrease, whereas those of species belonging to the phylum Proteobacteria increase [8]. Specific changes in gut microbial populations increase susceptibility to C difficile colonization [7]. Therefore, factors resulting in gut microbiota disruption, such as advanced age, hospitalization, antibiotic exposure, proton pump inhibitor (PPI) use, immunosuppression, and comorbidities (inflammatory bowel diseases, etc), are recognized as risk factors for CDI [6, 7].

Current clinical practice guidelines recommend vancomycin as the first-line option for pediatric CDI and fidaxomicin for adult CDI [2, 9]. However, conventional antibiotic treatment is limited by a high risk of recurrence, which is defined as an episode of symptom onset and positive assay result following an episode with positive assay result in the previous 2–8 weeks [2]. Recurrent CDI (rCDI) occurs in 15%–30% of patients who respond to first-round antibiotic treatment, and the subsequent recurrence rate can reach 45%–65% after the first and second recurrences [10]. Our previous study revealed that rCDI occurred in 48.53% (33/68) and 46.33% (19/41) of pediatric CDI patients treated with metronidazole and vancomycin, respectively [11]. Fecal microbiota transplantation (FMT) has been used to manage rCDI, with high cure rates in both adults and children [12–15].

To further explore the roles of the gut microbiota in primary and rCDI, several studies have compared the differences in the gut microbiota composition at the initial CDI diagnosis between individuals with and without recurrence. Khanna et al reported that, compared with patients without recurrence, patients with rCDI presented significant increases in Veillonella, Enterobacteriaceae, streptococci, Parabacteroides, and Lachnospiraceae before treatment [16]. Pakpour et al identified Veillonella dispar as a candidate organism for reducing risk of CDI recurrence [17]. Seekatz et al reported a less diverse microbial community in patients with rCDI [18]. Furthermore, a recent study conducted by Dawkins et al revealed that gut metabolites can predict disease recurrence at 1 week and 2 weeks posttreatment for primary CDI [19]. Taken together, these studies indicate that changes in the gut microbiota composition and metabolites are involved in rCDI and that specific microbiota-derived characteristics may be used to predict disease recurrence in adults. However, studies investigating the roles of the gut microbiota and metabolites in predicting rCDI in pediatric patients are still limited.

In this study, we aimed to investigate the characteristics of the fecal microbiota and metabolites in a group of pediatric patients at the initial diagnosis of CDI and to explore their relationships with disease recurrence.

MATERIALS AND METHODS

Study Design

A total of 84 pediatric patients with CDI were prospectively recruited to the study cohort from the Department of Gastroenterology, Hepatology and Nutrition of Shanghai Children's Hospital between September 2019 and August 2022. The inclusion criteria were children aged <16 years and diagnosed with primary (no episodes of CDI within the past 6 months), uncomplicated CDI by a positive C difficile test with either enzyme immunoassay toxin or polymerase chain reaction (PCR) and who were being treated for clinical symptoms. Patients with comorbidities of other gastrointestinal diseases, inherited or acquired immunodeficiencies, hepatic or renal dysfunction, hematologic malignancy, allergy, or fulminant colitis were excluded. Participant demographic and clinical data were extracted from the electronic medical records and systematically analyzed. Stool samples were collected at the time of initial diagnosis before antibiotic treatment for CDI. All the stool samples were stored at −80°C until bacterial 16S ribosomal RNA (rRNA) gene sequencing and fecal metabolite quantification were performed.

This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by the Ethical Review Board of the Shanghai Children's Hospital (2022R043-E01). Written informed consent was obtained from the parents or legal guardians of all the participants.

DNA Extraction and 16S rRNA Sequencing

Total fecal DNA was extracted from 250 μL of each fecal sample using a QIAamp DNA Stool Mini Kit (Qiagen, Germany) following the manufacturer's protocol. The eluted DNA was quantified using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific) and agarose gel electrophoresis. The isolated DNA was used for bacterial 16S rRNA V3–V4 hypervariable region amplification using the primer set 341F 5′-CCTACGGGAGGCAGCAG-3′/806R 5′-GGACTACHVGGGTWTCTAAT-3′ as previously described [20]. The purified PCR amplicons were subjected to barcoded library construction using the TruSeq Nano DNA LT Library Prep Kit (Illumina) according to the manufacturer's instructions. Library quality was assessed and quantified on an Agilent 2100 Bioanalyzer (Agilent Technologies). Each library was diluted to a final concentration of 12.5 nM and sequenced on an Illumina NovaSeq platform (Illumina).

Sequencing Data Processing and Analysis

The raw sequence data were checked using FastQC (Babraham Bioinformatics), and forward and reverse reads were joined and quality trimmed using fastq_maxee 1.0. Paired-end reads were de-noised using DADA2 software and further analyzed using the QIIME2 platform (v.2020.2) [21, 22]. The taxonomic assignment of amplicon sequence variants (ASVs) was classified using the Greengenes (v.13.8) reference database implemented in QIIME2 (v.2020.2). To reduce the effect of spurious sequences, ASVs with <0.005% of the total number of sequences were removed. The alpha diversity was evaluated using the observed species, abundance-based coverage estimator (ACE), and Chao1, Shannon, and Simpson indices. The beta diversity was calculated using principal coordinate analysis (PCoA) of the Bray-Curtis distance based on the ASV abundance. The significant differences in alpha diversity and ASV abundances between groups were analyzed using the Wilcoxon rank-sum test, with P values adjusted for the false discovery rate. Permutational multivariate analysis of variance was used to estimate the significant differences in the gut microbiota structure. For the relative abundance ratio of bacteria at the phylum, class, order, and family levels, rare taxa of ASVs present with <10 counts and in <10% of the fecal samples were removed prior to differential abundance analysis. The linear discriminant analysis (LDA) effect size (LEfSe) [23] method was used to assess the significance and biological relevance of the enriched taxa in different groups.

Quantification of Fecal Metabolites

Targeted metabolomic profiling was performed to quantify fecal metabolites as previously described [24]. In brief, 5 mg of each lyophilized fecal sample was dissolved in 25 µL of sterile water and homogenized with zirconium oxide beads for 3 minutes, after which 120 µL of methanol containing internal standards was added. After homogenization for 3 minutes, the sample was centrifuged at 18 000g at 4°C for 20 minutes. A 20-µL supernatant aliquot of each sample was transferred to a 96-well plate, and 20 µL of freshly prepared derivative reagents was added to each well. After 60 minutes of incubation at 30°C for derivatization, the sample was diluted with 330 μL of ice-cold 50% methanol solution. The plate was then incubated at −20°C for 20 minutes and subsequently centrifuged at 4000g at 4°C for 30 minutes. A volume of 135 μL of supernatant per sample was transferred to another 96-well plate with 10 μL of internal standards in each well. A 5-μL aliquot of each sample was injected to quantify fecal metabolites using ultraperformance liquid chromatography coupled with a tandem mass spectrometry (UPLC-MS/MS) system (ACQUITY UPLC-Xevo TQ-S) at a flow rate of 0.4 mL/minute.

The UPLC-MS/MS raw data were processed using TMBQ software (Metabo-Profile, China) to perform peak integration, calibration, and quantitation for each metabolite. iMAP (Metabo-Profile, China) was used for statistical analysis and feature selection of the metabolomic data, including principal component analysis and partial least squares discriminant analysis (PLS-DA). PLS-DA was used to analyze the variable importance in projection (VIP) values of each metabolite between groups. The Wilcoxon rank-sum test with P values adjusted with the Bonferroni correction procedure was performed for univariate statistical analysis. Differentially abundant metabolites with VIP scores of PLS-DA >1, P < .05, and log2 (fold change) >1 were selected. Volcano plots and heatmaps were generated via R software.

Statistical Analyses

The demographic and clinical data are presented as numbers (percentages) for dichotomous variables and medians with interquartile ranges (IQRs [25th to 75th percentiles]) for continuous variables. The χ2 test and the Mann-Whitney U test were used for statistical analysis of categorical and continuous variables, respectively. Statistical analysis was performed using SPSS software (version 25.0), and a P value of <.05 was considered statistically significant.

RESULTS

Participant Characteristics

The median age of the 84 enrolled pediatric patients with primary CDI was 4.5 years (IQR, 2.4–7.8 years), and 49 (58.3%) were boys (Supplementary Table 1). A history of exposure to antibiotics and PPI use occurred in 64 (76.2%) and 21 (25.0%) patients, respectively. The main clinical symptoms were diarrhea (60/84 [17.4%]), abdominal pain (42/84 [50.0%]), and hematochezia (35/84 [41.7%]). For initial antibiotic treatment, 46 (54.8%) patients received metronidazole, and 38 (45.2%) patients received vancomycin (Supplementary Table 1). As shown in Table 1, 58 (69.0%) patients were cured by 1 round of conventional antibiotic treatment, whereas 26 (31.0%) patients experienced disease recurrence. The ratios of female sex and PPI exposure were greater in recurrent patients than in nonrecurrent patients. There were no significant differences in clinical symptoms between recurrent and nonrecurrent patients. In addition, the disease recurrence rate was comparable in patients treated with either metronidazole or vancomycin.

Table 1. Demographic and Clinical Data of Participants With Clostridioides difficile Infection

Characteristic	Nonrecurrers (n = 58)	Recurrers (n = 26)	P Value	
Sex, male, No. (%)	38.0 (65.5)	11.0 (42.3)	.05	
Age, y, median (IQR)	4.7 (2.5–7.5)	4.0 (2.0–7.7)	.58	
BMI, kg/m2, median (IQR)	14.8 (13.1–16.8)	14.3 (13.1–16.4)	.77	
Exposure history, No. (%)				
 Antibiotics	44.0 (75.9)	20.0 (76.9)	.91	
 PPI	10.0 (17.2)	11.0 (42.3)	.02	
Symptoms, No. (%)				
 Fever	11.0 (20.0)	3.0 (11.5)	.40	
 Vomiting	11.0 (20.0)	8.0 (30.8)	.24	
 Diarrhea	39.0 (67.2)	21.0 (80.8)	.21	
 Abdominal pain	28.0 (48.3)	14.0 (53.8)	.64	
 Hematochezia	21.0 (36.2)	14.0 (53.8)	.13	
 Pseudomembrane	9.0 (15.5)	7.0 (26.9)	.22	
Initial treatment antibiotic used, No. (%)				
 Metronidazole	35.0 (60.3)	11.0 (42.3)	.23	
 Vancomycin	23.0 (39.7)	15.0 (57.7)	.23	
Abbreviations: BMI, body mass index; IQR, interquartile range; PPI, proton pump inhibitor.

Gut Microbiota Analysis

The average number of raw reads per sample generated by sequencing was 95 720, ranging from 83 354 to 111 770. A total of 6 534 142 high-quality bacterial 16S rRNA gene sequences spanning 2280 ASVs were obtained after sequence processing and filtering. Alpha diversity analysis revealed that the richness calculated for the observed species (P < .001), ACE (P < .01), and Chao1 indices (P < .001) and the Shannon (P < .01) and Simpson indices (P < .05) were significantly lower in the recurrent samples than in the nonrecurrent samples (Figure 1A–E). PCoA of beta diversity using the Bray-Curtis distance revealed that the microbial community of the recurrent group was separated from that of the nonrecurrent group (Figure 1F). The bacterial microbiota was dominated by the phyla Firmicutes, Proteobacteria, Actinobacteria, Bacteroidetes, and Verrucomicrobia in both groups (Figure 2A).

Figure 1. Fecal microbiota biodiversity of children with Clostridioides difficile infection (CDI). A–E, Alpha diversity measured in the observed species, abundance-based coverage estimator (ACE), Chao1, Shannon, and Simpson indices. Signiﬁcance was determined using the Wilcoxon rank-sum test, with *P < .05, **P < .01, and ***P < .001. F, Beta diversity analyzed by principal coordinate analysis (PCoA) of the Bray-Curtis distance on amplicon sequence variants (ASVs). The permutational multivariate analysis of variance (PERMANOVA) method was used to compare the differences between groups. NR indicates nonrecurrent; R indicates recurrent.

Figure 2. Fecal microbiota composition of children with Clostridioides difficile infection (CDI). A–D, Relative abundance ratios of bacteria at the phylum, class, order, and family levels between the nonrecurrent (NR) and recurrent (R) groups. The top 5 abundant phyla and the top 15 abundant classes, orders, and families are indicated. E, Genera with significant differences between the NR and R groups. The relative abundances are shown in log10 values. Statistical significance was assessed via the Wilcoxon rank-sum test, with *P < .05, **P < .01, and ***P < .001.

Taxonomic proﬁles revealed the relative abundances of the phyla, class, order, and family between the 2 groups. Phyla Firmicutes, class Clostridia, order Clostridiales, and family Lachnospiraceae were mostly decreased, whereas the abundances of the phyla Proteobacteria, class Gammaproteobacteria, order Enterobacterales, and family Enterobacteriaceae were mostly increased in the recurrent group compared with those in the nonrecurrent group (Figure 2A–D). Intergroup comparisons revealed significantly lower relative abundances of the genera Blautia, Clostridium, Coprococcus, Dorea, Eggerthella, Eubacterium, and Subdoligranulum and significantly greater abundances of the genus Ethanoligenens and unidentified Synergistaceae in the recurrent versus nonrecurrent group (Figure 2E, Supplementary Table 2).

LEfSe analysis identified multiple taxa at different levels that were differentially abundant between recurrent and nonrecurrent patients. The class Clostridia, order Clostridiales; family Coriobacteriaceae; and genera Adlercreutzia, Eggerthella, Parabacteroides, Blautia, and Clostridium were enriched in nonrecurrers, whereas the classes Chthonomonadetes and Bacilli, order Bacillales, family Helicobacteraceae, and genera Ethanoligenens and Helicobacter were enriched in recurrers (Supplementary Figure 1A). Specifically, the classes Bacilli, Clostridia, and Erysipelotrichi; orders Clostridiales, Lactobacillales, and Erysipelotrichales; families Prophyromonadaceae, Erysipelotrichaceae, Rumincoccaceae, and Lachnospiraceae; and genera Parabacteroides, Coprococcus, Dialister, Clostridium, Faecalibacterium, Blautia, Eubacterium, and Ruminococcus were differentially abundant between the 2 groups, with LDA scores >3.0 (Supplementary Figure 1B).

Metabolic Profiling

A total of 250 metabolites were detected and quantified by targeted metabolomics analysis in all the fecal samples. Amino acids, short-chain fatty acids (SCFAs), carbohydrates, bile acids (BAs), and organic acids were the dominant fecal metabolites in both groups (Figure 3). Different distribution patterns of metabolites between recurrent and nonrecurrent patients were observed using both PCoA (Figure 4A) and PLS-DA (Figure 4B). Among the 19 metabolites that were the most discriminative (VIP score >1.5) for classification between the 2 groups, the metabolites of the recurrent group presented higher fecal levels of 2-hydroxyglutaric acid, lactic acid, arachidonic acid, acetylcarnitine, hydroxylisovalerylcarnitine, isovalerylcarnitine, hexanylcarnitine, threonic acid, propionylcarnitine, creatine, 2-methylbutyroylcarnitine, linoleic acid, and erythronic acid, and lower levels of isovaleric acid, thiophenecarboxylic acid, tauroursodeoxycholic acid, ethylmethylacetic acid, butyric acid, and nicotinic acid (Figure 4C).

Figure 3. Relative abundance of each metabolite class in fecal samples from patients with nonrecurrent and recurrent Clostridioides difficile infection. Abbreviations: NR, nonrecurrent; R, recurrent; SCFA, short-chain fatty acid.

Figure 4. Fecal metabolome of children with Clostridioides difficile infection (CDI). A, Principal coordinate analysis plot of fecal metabolites of patients with nonrecurrent (NR) and recurrent (R) CDI according to the Bray-Curtis distance. The permutational multivariate analysis of variance method was used to compare the differences between groups. B, Partial least squares discriminant analysis models discriminated fecal metabolites between the NR and R groups. C, The top metabolites ranked by variable importance in projection scores. Abbreviations: NR, nonrecurrent; PCoA, principal coordinate analysis; PERMANOVA, permutational multivariate analysis of variance; PLS-DA, partial least squares discriminant analysis; R, recurrent; THCA, thiophenecarboxylic acid; TUDCA, tauroursodeoxycholic acid; VIP, variable importance in projection.

As shown in Figure 4 and Supplementary Table 3, the concentrations of 35 unique fecal metabolites differed significantly between the recurrent and nonrecurrent samples. Among them, the concentrations of lithocholic acid (LCA), 12-ketoLCA, trihydroxycholestanoic acid, butyric acid, deoxycholic acid (DCA), ethylmethylacetic acid, glutaric acid, isobutyric acid, isoLCA, isovaleric acid, methylsuccinic acid, N-methylnicotinamide, nicotinic acid, phenylacetic acid, pipecolic acid, and valeric acid decreased, whereas the concentrations of 2-hydroxyglutaric acid, 2-methylbutyroylcarnitine, 3-hydroxylisovalerylcarnitine, acetylcarnitine, α-linolenic acid, arachidonic acid, butyrylcarnitine, carnitine, creatine, docosahexaenoic acid, eicosapentaenoic acid, erythronic acid, γ-linolenic acid, hexanylcarnitine, isovalerylcarnitine, lactic acid, linoleic acid, propionylcarnitine, and threonic acid increased significantly in patients with recurrence compared with nonrecurrence (Supplementary Figure 2).

Covariance Between Gut Bacteria and Metabolites

Associations between the differential gut bacteria and metabolites in fecal samples were coupled by Spearman rank correlations. Strong correlations between 22 bacterial genera and 35 metabolites were observed in nonrecurrent and recurrent samples (Figure 5). A total of 19 metabolites that increased in the recurrent group were positively correlated with the enriched genera in the recurrent group and negatively correlated with the genera that were enriched in the nonrecurrent group. For example, the decreased levels of butyric acid, isobutyric acid, isovaleric acid, and valeric acid in the recurrent group were associated with decreased abundances of SCFA producers, such as Faecalibacterium, Clostridium, Butyricicoccus, Roseburia, and Ruminococcus. Conversely, the enriched SCFA producers were positively correlated with high concentrations of fecal SCFAs in nonrecurrent patients. Furthermore, the genera enriched in the recurrent group, including Helicobacter, Ethanoligenens, and Streptococcus, were negatively correlated with the levels of isoLCA, 12-ketoLCA, and DCA.

Figure 5. Heatmap of the correlation matrix of different fecal bacterial genera and metabolites between patients with nonrecurrent and recurrent Clostridioides difficile infection. *P < .05, **P < .01, ***P < .001. Abbreviations: 3-DHCA, trihydroxycholestanoic acid; DCA, deoxycholic acid; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; LCA, lithocholic acid; NR, nonrecurrent; R, recurrent.

DISCUSSION

Owing to the increase in community-acquired infections and the emergence of hypervirulent C difficile strains, the incidence and severity of pediatric CDI have increased in recent decades [2, 25–27]. Although conventional antibiotic-based treatments are generally effective for managing pediatric CDI, a high risk of disease recurrence is challenging [2]. The treatment of rCDI is more difficult than the initial episode because of the significantly increased rate of further recurrence [10]. Identification of risk factors related to rCDI is vital for early detection, treatment, and prevention of disease recurrence. Early alternative therapies, such as FMT, could be used to treat CDI patients who are likely to experience recurrence.

To our knowledge, our study represents the first prospective study focused on the role of the gut microbiota in the occurrence of rCDI in pediatric patients, employing both fecal microbiome sequencing and targeted metabolomics analyses. With respect to the results of the fecal microbiome analysis, we reported a significantly lower alpha diversity of the fecal microbiota in pediatric CDI patients than in nonrecurrent patients prior to antibiotic treatment, which was consistent with the findings of previous studies conducted in adult CDI patients [16–19]. The gut microbiota signatures at the initial diagnosis were characterized by increasing abundances of taxa belonging to the phyla Proteobacteria and Bacteroidetes but decreasing abundances of species of the phyla Firmicutes and Actinobacteria. Specifically, recurrent patients had significantly greater abundances of Enterobacteriaceae, Streptococcaceae, Veillonellaceae, Porphyromonadaceae, Prevotellaceae, Shigella, Streptococcus, Veillonella, Prevotella, and Dysgonomonas than nonrecurrent patients. Similarly, Enterobacteriaceae, Veillonella, and streptococci abundances were significantly increased in pre-antibiotic-treated fecal samples from adult CDI patients with recurrence [16]. The persistent expansion of Enterobacteriaceae may increase susceptibility to C difficile colonization [28]. However, Pakpour et al reported that Veillonella dispar is a candidate organism for preventing CDI recurrence, which was contrary to our study [17]. A study in 2023 indicated that Fusobacterium increased during rCDI, whereas Collinsella, Senegalimassilia, Prevotella, and Ruminococcus decreased, findings that were partly similar to our data [29]. We found that the genus Parabacteroides was differentially abundant between the 2 groups, with LDA scores >3.0, which means that it may be a key microbiota for preventing rCDI. Parabacteroides distasonis, which can produce succinate, has been reported to be associated with CDI. Many studies have shown that Parabacteroides distasonis is inversely related to rheumatoid arthritis, obesity, and diabetes mellitus [30, 31]. It may also play an important role in C difficile expansion [16, 32]. In addition, the decreased abundances of Lachnospiraceae, Blautia, and Faecalibacterium reported in adults with recurrent CDI were also decreased in our pediatric patients [19]. Future studies are needed to investigate the mechanisms of these bacteria that may be involved in rCDI after successful antibiotic treatment.

Gut-associated metabolites are key mediators of microbiota‒host interactions. A variety of host- and microbial-derived metabolites play important roles in C difficile pathogenesis, including BAs, SCFAs, organic acids, and amino acids [33, 34]. Certain BAs have been shown to either promote or inhibit C difficile germination or affect C difficile growth and colonization [33, 34]. For example, both taurocholic acid and cholic acid initiate C difficile spore germination in the small intestine, whereas chenodeoxycholic acid suppresses the germination of C difficile spores [35–37]. IsoalloLCA inhibited C difficile growth both in vitro and in vivo [34]. DCA is known to decrease the sporulation efficiency of C difficile by downregulating the expression of the master regulator of sporulation, Spo0A [35, 37]. DCA can inhibit C difficile growth and repress toxin activity [38]. SCFAs and organic acids are small molecules derived from the bacterial fermentation of dietary carbohydrates. SCFAs of acetate and propionate had inhibitory effects on toxin production and the growth of C difficile [39]. Furthermore, several organic acids and amino acids have been reported to be involved in CDI, such as succinate and sialic acid, which promote C difficile expansion [40, 41], whereas glycine, proline, and cysteine inhibit the toxin yield of C difficile [42–44]. Our data revealed that the fecal levels of several BAs (12-ketoLCA, isoLCA, LCA, and DCA), SCFAs (butyric acid, glutaric acid, isobutyric acid, isovaleric acid, and valeric acid), and organic acids (methylsuccinic acid, nicotinic acid, phenylacetic acid, and amino acids [pipecolic acid]) were significantly decreased, suggesting that these metabolites may be involved in the disease recurrence of pediatric CDI. Nevertheless, future studies are warranted to further investigate the mechanisms of interactions between gut-associated metabolites and C difficile pathogenesis.

There are several limitations to our study. First, this was a single-center study with a limited number of enrolled participants. Studies with large populations of pediatric CDI patients are needed. Second, only fecal samples before treatment were collected for gut microbiota and metabolite analysis. Longitudinal studies dedicated to the dynamics of the microbiome and metabolome are needed to advance our understanding of the interactions between the gut microbiota/metabolites and C difficile pathogenesis. Third, we only evaluated bacterial composition changes using 16S rRNA sequencing; other microbiota components, such as viruses and fungi, need to be investigated in future studies. Finally, the mechanisms of the potential gut microbiota and metabolite changes involved in CDI recurrence are underexplored in the current study.

In summary, our data indicate that the disturbed gut microbiota in pediatric CDI patients may contribute to disease recurrence by modulating microbial metabolism characterized by altered BA profiles and reduced SCFA concentrations. The gut microbiota associated with BA and SCFA signatures may be used to predict disease recurrence in pediatric CDI patients. In the future, more studies on the microbiota and BA and SCFA in children with CDI should be conducted to determine their beneficial effects on the development of methods for diagnosing and treating CDI recurrence.

Supplementary Material

ofae506_Supplementary_Data

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Author contributions. Y. W. and T. Z. conceived the study. X. L. and F. X. performed the experiment. Y. W., X. L., and F. X. analyzed the data and created figures. L. Y., X. W., and Y. X. contributed to data curation, formal analysis, and software. D. L. contributed to stool sample collection. Y. W. and X. L. interpreted the data and wrote the manuscript. T. Z. edited the manuscript. All authors read and approved the final manuscript.

Data availability. Raw sequencing data have been deposited in the National Center for Biotechnology Information Sequence Read Archive database under the BioProject ID PRJNA1020794 and BioSamples ID SUB13861569.

Financial support. This work was supported by the National Natural Science Foundation of China (grant numbers 32370951 and 81900472); the Natural Science Foundation of Shanghai (grant number 22ZR1451800); the Shanghai Municipal Health Commission (grant number 20214Y0349); the training program for young talents of Shanghai Children's Hospital (grant number 2021YQ03); and 2023 Xinyi Digestive Disease Research Foundation, School of Digestive Science, Shanghai Jiao Tong University School of Medicine (grant number KY-2023-08-01).
==== Refs
References

1 Rupnik  M, Wilcox  MH, Gerding  DN. Clostridium difficile infection: new developments in epidemiology and pathogenesis. Nat Rev Microbiol  2009; 7 :526–36.19528959
2 McDonald  LC, Gerding  DN, Johnson  S, et al  Clinical practice guidelines for Clostridium difficile infection in adults and children: 2017 update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA). Clin Infect Dis  2018; 66 :e1–48.29462280
3 Buddle  JE, Fagan  RP.  Pathogenicity and virulence of Clostridioides difficile. Virulence  2023; 14 :2150452.36419222
4 De Roo  AC, Regenbogen  SE. Clostridium difficile infection: an epidemiology update. Clin Colon Rectal Surg  2020; 33 :49–57.32104156
5 Balsells  E, Shi  T, Leese  C, et al  Global burden of Clostridium difficile infections: a systematic review and meta-analysis. J Glob Health  2019; 9 :010407.30603078
6 Kelly  CR, Fischer  M, Allegretti  JR, et al  ACG clinical guidelines: prevention, diagnosis, and treatment of Clostridioides difficile infections. Am J Gastroenterol  2021; 116 :1124–47.34003176
7 Samarkos  M, Mastrogianni  E, Kampouropoulou  O. The role of gut microbiota in Clostridium difficile infection. Eur J Intern Med  2018; 50 :28–32.29428498
8 Gonzales-Luna  AJ, Carlson  TJ, Garey  KW. Gut microbiota changes associated with Clostridioides difficile infection and its various treatment strategies. Gut Microbes  2023; 15 :2223345.37318134
9 Johnson  S, Lavergne  V, Skinner  AM, et al  Clinical practice guideline by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA): 2021 focused update guidelines on management of Clostridioides difficile infection in adults. Clin Infect Dis  2021; 73 :e1029–44.34164674
10 Song  JH, Kim  YS. Recurrent Clostridium difficile infection: risk factors, treatment, and prevention. Gut Liver  2019; 13 :16–24.30400734
11 Li  X, Xiao  F, Li  Y, et al  Characteristics and management of children with Clostridioides difficile infection at a tertiary pediatric hospital in China. Braz J Infect Dis  2022; 26 :102380.35753395
12 Cammarota  G, Ianiro  G, Tilg  H, et al  European consensus conference on faecal microbiota transplantation in clinical practice. Gut  2017; 66 :569–80.28087657
13 Jorgensen  SMD, Hansen  MM, Erikstrup  C, Dahlerup  JF, Hvas  CL. Faecal microbiota transplantation: establishment of a clinical application framework. Eur J Gastroenterol Hepatol  2017; 29 :e36–45.28863010
14 Drekonja  D, Reich  J, Gezahegn  S, et al  Fecal microbiota transplantation for Clostridium difficile infection: a systematic review. Ann Intern Med  2015; 162 :630–8.25938992
15 Li  X, Gao  X, Hu  H, et al  Clinical efficacy and microbiome changes following fecal microbiota transplantation in children with recurrent Clostridium difficile infection. Front Microbiol  2018; 9 :2622.30450088
16 Khanna  S, Montassier  E, Schmidt  B, et al  Gut microbiome predictors of treatment response and recurrence in primary Clostridium difficile infection. Aliment Pharmacol Ther  2016; 44 :715–27.27481036
17 Pakpour  S, Bhanvadia  A, Zhu  R, et al  Identifying predictive features of Clostridium difficile infection recurrence before, during, and after primary antibiotic treatment. Microbiome  2017; 5 :148.29132405
18 Seekatz  AM, Rao  K, Santhosh  K, Young  VB. Dynamics of the fecal microbiome in patients with recurrent and nonrecurrent Clostridium difficile infection. Genome Med  2016; 8 :47.27121861
19 Dawkins  JJ, Allegretti  JR, Gibson  TE, et al  Gut metabolites predict Clostridioides difficile recurrence. Microbiome  2022; 10 :87.35681218
20 Wang  Y, Gao  X, Zhang  X, et al  Microbial and metabolic features associated with outcome of infliximab therapy in pediatric Crohn's disease. Gut Microbes  2021; 13 :1–15.
21 Bolyen  E, Rideout  JR, Dillon  MR, et al  Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol  2019; 37 :852–7.31341288
22 Callahan  BJ, McMurdie  PJ, Rosen  MJ, Han  AW, Johnson  AJ, Holmes  SP. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods  2016; 13 :581–3.27214047
23 Segata  N, Izard  J, Waldron  L, et al  Metagenomic biomarker discovery and explanation. Genome Biol  2011; 12 :R60.21702898
24 Xie  G, Wang  L, Chen  T, et al  A metabolite array technology for precision medicine. Anal Chem  2021; 93 :5709–17.33797874
25 Alvarez  AM, Rathore  MH. Clostridium difficile infection in children. Adv Pediatr  2019; 66 :263–80.31230698
26 Malmqvist  L, Ullberg  M, Hed Myrberg  I, Nilsson  A. Clostridium difficile infection in children: epidemiology and trend in a Swedish tertiary care hospital. Pediatr Infect Dis J  2019; 38 :1208–13.31738336
27 Miranda-Katz  M, Parmar  D, Dang  R, Alabaster  A, Greenhow  TL. Epidemiology and risk factors for community associated Clostridioides difficile in children. J Pediatr  2020; 221 :99–106.32171559
28 Schubert  AM, Rogers  MA, Ring  C, et al  Microbiome data distinguish patients with Clostridium difficile infection and non-C. difficile-associated diarrhea from healthy controls. mBio  2014; 5 :e01021–01014.24803517
29 Vazquez-Cuesta  S, Garcia  NL, Fernandez  AI, et al  Microbiome profile and calprotectin levels as markers of risk of recurrent Clostridioides difficile infection. Front Cell Infect Microbiol  2023; 13 :1237500.37780848
30 Sun  H, Guo  Y, Wang  H, et al  Gut commensal Parabacteroides distasonis alleviates inflammatory arthritis. Gut  2023; 72 :1664–77.36604114
31 Wang  K, Liao  M, Zhou  N, et al  Parabacteroides distasonis alleviates obesity and metabolic dysfunctions via production of succinate and secondary bile acids. Cell Rep  2019; 26 :222–35.e5.30605678
32 Lawley  TD, Clare  S, Walker  AW, et al  Targeted restoration of the intestinal microbiota with a simple, defined bacteriotherapy resolves relapsing Clostridium difficile disease in mice. PLoS Pathog  2012; 8 :e1002995.23133377
33 Aguirre  AM, Sorg  JA. Gut associated metabolites and their roles in Clostridioides difficile pathogenesis. Gut Microbes  2022; 14 :2094672.35793402
34 Gao  Y, Ma  L, Su  J. Host and microbial-derived metabolites for Clostridioides difficile infection: contributions, mechanisms and potential applications. Microbiol Res  2022; 263 :127113.35841835
35 Sorg  JA, Sonenshein  AL. Bile salts and glycine as cogerminants for Clostridium difficile spores. J Bacteriol  2008; 190 :2505–12.18245298
36 Sorg  JA, Sonenshein  AL. Chenodeoxycholate is an inhibitor of Clostridium difficile spore germination. J Bacteriol  2009; 191 :1115–7.19060152
37 Sorg  JA, Sonenshein  AL. Inhibiting the initiation of Clostridium difficile spore germination using analogs of chenodeoxycholic acid, a bile acid. J Bacteriol  2010; 192 :4983–90.20675492
38 Thanissery  R, Winston  JA, Theriot  CM. Inhibition of spore germination, growth, and toxin activity of clinically relevant C. difficile strains by gut microbiota derived secondary bile acids. Anaerobe  2017; 45 :86–100.28279860
39 Fachi  JL, Secca  C, Rodrigues  PB, et al  Acetate coordinates neutrophil and ILC3 responses against C. difficile through FFAR2. J Exp Med  2020; 217 :jem.20190489.31876919
40 Ferreyra  JA, Wu  KJ, Hryckowian  AJ, Bouley  DM, Weimer  BC, Sonnenburg  JL. Gut microbiota-produced succinate promotes C. difficile infection after antibiotic treatment or motility disturbance. Cell Host Microbe  2014; 16 :770–7.25498344
41 Ng  KM, Ferreyra  JA, Higginbottom  SK, et al  Microbiota-liberated host sugars facilitate post-antibiotic expansion of enteric pathogens. Nature  2013; 502 :96–9.23995682
42 Karlsson  S, Burman  LG, Akerlund  T. Suppression of toxin production in Clostridium difficile VPI 10463 by amino acids. Microbiology (Reading)  1999; 145 (Pt 7 ):1683–93.10439407
43 Karlsson  S, Burman  LG, Akerlund  T. Induction of toxins in Clostridium difficile is associated with dramatic changes of its metabolism. Microbiology (Reading)  2008; 154 :3430–6.18957596
44 Karlsson  S, Lindberg  A, Norin  E, Burman  LG, Akerlund  T. Toxins, butyric acid, and other short-chain fatty acids are coordinately expressed and down-regulated by cysteine in Clostridium difficile. Infect Immun  2000; 68 :5881–8.10992498
