Article Text
PDF
Statistics from Altmetric.com
We read with great interest the article by Wu et al1 showing that the impact of diet on the composition of the gut microbiota was milder than its impact on metabolites produced by the gut microbiota, in healthy humans. Enteral nutrition is an effective treatment to manage active Crohn's disease (CD) in children.2 CD is known to be associated with gut microbial dysbiosis that is characterised by decreased diversity and imbalances in the intestinal microbiome. Recent evidence indicates that exclusive enteral nutrition (EEN) may affect gut microbiota in children with CD,3 but it remains elusive whether the aforementioned changes may occur independently of disease flares. We thus studied the impact of either EEN with Modulen (EXCL-MOD) or partial enteral nutrition (PEN) with Modulen (MOD) IBD on the bacterial composition of the gut microbiota from children with CD during either remission or active phase.
In this single-centre, prospective, observational cohort study, 34 children followed at Nancy University Children Hospital were included. Clinical and biological data were recorded. Remission (Rem) or relapse (Rel) status was defined according to faecal calprotectin levels using 250 µg/g as a cut-off. Median follow-up was 7 months. Median age at inclusion was 14.8 years (range, 6.5–21) and 38.2% (n=13) were female. The faecal microbiota of all included patients was analysed by MiSeq sequencing of the 16S rRNA gene. A total of 1 593 260 paired-end sequence reads were obtained. From these reads, an average of 25 085 (range 16 541–35 011) reads per sample was kept for further analyses after filtration with usearch (PMID: 20709691)4 for a minimum length of 250 bp, and for an E_max score of 0.25. Sequences were clustered into phylotypes using uclust (similarity threshold 0.97) and representative phylotypes were assigned to their closest taxa, from phylum to isolates levels, based on the RDP database. For the four children with newly diagnosed CD belonging to the EEN group (EXCL-MOD), faecal samples were collected immediately prior to EEN initiation and also 2 weeks and 6 weeks after the start of EEN. For children who were treated with PEN (MOD) or without receiving enteral nutrition (NO-MOD) (MOD-Rem, n=3; NO-MOD-Rem, n=12; MOD-Rel, n=5; and NO-MOD-Rel, n=10), only one faecal sample was collected.
Interestingly, interclass PCA analysis of bacterial genera revealed that children in the MOD-Rem group exhibited a specific microbiota with a significant decrease in Dialister (mainly D. invisus), Blautia, unclassified Ruminococcaceae and Coprococcus as compared with the NO-MOD-Rem group (figure 1A). In the EXCL-MOD group, a decrease in genera from the Proteobacteria phylum (Escherichia-Shigella p<0.1 and Sutterella p<0.05) was observed at weeks 2 and 6, whereas Alistipes proportions significantly increased over time (figure 1B).
Effects of enteral nutrition with Modulen IBD on the gut microbiota. (A) Interclass PCA analysis of bacterial genera distribution. Each black dot stands for one sample and the ellipses group the three different clinical groups, that is, MOD (partial enteral nutrition with Modulen), NO-MOD (No Modulen) and EXCL-MOD (exclusive enteral nutrition with Modulen). The clustering based on microbial distribution is significant (Monte-Carlo p value: 0.029 based on 10 000 replicates). Component 1=7.6%, Component 2=3.8%. (B) Over time modifications of specific bacterial genera during exclusive Modulen treatment (n=3 patients). W0, immediately before starting Modulen; W2, two weeks after starting Modulen; W6, 6 weeks after starting Modulen.
Importantly, Actinobacteria and Clostridia classes were negatively correlated with faecal calprotectin levels (r=−0.41; p<0.01) and Harvey Bradshaw scores (r=−0.38; p=0.01). β-Proteobacteria was positively correlated with faecal calprotectin levels (r=0.31; p=0.05) and γ-Proteobacteria was positively correlated with Harvey Bradshaw scores (r=0.38; p=0.01) in all groups (figure 2).
Correlation between dominant microbial genera and clinical parameters or faecal calprotectin levels. A Spearman's correlation was applied and only r>0.3 and r<−0.3 are shown in the heatmap. All correlations have a p value ≤0.05 except unclassified Enterococcaceae, Bilophila, unclassified Enterobacteriaceae, Papillibacter, unclassified Clostridiales, Acetanaerobacterium and Acinetobacter. HB, Harvey-Bradshaw score.
In conclusion, enteral polymeric diet with Modulen IBD, either exclusive or partial, had significant effects on gut microbiota in children with CD. Changes in the microbiota were correlated with clinical and biological data and occurred early after the start of EEN. Further investigations are needed to confirm our findings.
Footnotes
Contributors CG-C, research, data collection, analysis of data, drafting of the manuscript; PL, analysis and interpretation of data, drafting of the manuscript; AM, data collection and drafting of the manuscript; MC, analysis and interpretation of data, drafting of the manuscript; LP-B, analysis of data, drafting of the manuscript and study supervision..
Funding National grant from Picot (Lactalis group).
Competing interests LP-B: Consulting fees from Merck, Abbvie, Janssen, Genentech, Mitsubishi, Ferring, Norgine, Tillots, Vifor, Therakos, Pharmacosmos, Pilège, BMS, UCB-pharma, Hospira, Celltrion, Takeda, Biogaran, Boerhinger-Ingelheim, Lilly, Pfizer, HAC-Pharma, Index Pharmaceuticals, Amgen, Sandoz, Forward Pharma GmbH, Celgene. Lecture fees from Merck, Abbvie, Takeda, Janssen, Takeda, Ferring, Norgine, Tillots, Vifor, Therakos, Mitsubishi, HAC-pharma.
Provenance and peer review Not commissioned; externally peer reviewed.